Commit
•
40d58fc
1
Parent(s):
2216bb0
Upload 68 files
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- checkpoint-1148/config.json +36 -0
- checkpoint-1148/merges.txt +0 -0
- checkpoint-1148/model.safetensors +3 -0
- checkpoint-1148/optimizer.pt +3 -0
- checkpoint-1148/rng_state.pth +3 -0
- checkpoint-1148/scheduler.pt +3 -0
- checkpoint-1148/special_tokens_map.json +30 -0
- checkpoint-1148/tokenizer.json +0 -0
- checkpoint-1148/tokenizer_config.json +20 -0
- checkpoint-1148/trainer_state.json +831 -0
- checkpoint-1148/training_args.bin +3 -0
- checkpoint-1148/vocab.json +0 -0
- checkpoint-2296/config.json +36 -0
- checkpoint-2296/merges.txt +0 -0
- checkpoint-2296/model.safetensors +3 -0
- checkpoint-2296/optimizer.pt +3 -0
- checkpoint-2296/rng_state.pth +3 -0
- checkpoint-2296/scheduler.pt +3 -0
- checkpoint-2296/special_tokens_map.json +30 -0
- checkpoint-2296/tokenizer.json +0 -0
- checkpoint-2296/tokenizer_config.json +20 -0
- checkpoint-2296/trainer_state.json +1636 -0
- checkpoint-2296/training_args.bin +3 -0
- checkpoint-2296/vocab.json +0 -0
- checkpoint-3444/config.json +36 -0
- checkpoint-3444/merges.txt +0 -0
- checkpoint-3444/model.safetensors +3 -0
- checkpoint-3444/optimizer.pt +3 -0
- checkpoint-3444/rng_state.pth +3 -0
- checkpoint-3444/scheduler.pt +3 -0
- checkpoint-3444/special_tokens_map.json +30 -0
- checkpoint-3444/tokenizer.json +0 -0
- checkpoint-3444/tokenizer_config.json +20 -0
- checkpoint-3444/trainer_state.json +2441 -0
- checkpoint-3444/training_args.bin +3 -0
- checkpoint-3444/vocab.json +0 -0
- checkpoint-4592/config.json +35 -0
- checkpoint-4592/merges.txt +0 -0
- checkpoint-4592/model.safetensors +3 -0
- checkpoint-4592/optimizer.pt +3 -0
- checkpoint-4592/rng_state.pth +3 -0
- checkpoint-4592/scheduler.pt +3 -0
- checkpoint-4592/special_tokens_map.json +6 -0
- checkpoint-4592/tokenizer.json +0 -0
- checkpoint-4592/tokenizer_config.json +20 -0
- checkpoint-4592/trainer_state.json +3246 -0
- checkpoint-4592/training_args.bin +3 -0
- checkpoint-4592/vocab.json +0 -0
- checkpoint-5740/config.json +35 -0
- checkpoint-5740/merges.txt +0 -0
checkpoint-1148/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "C:/Users/wonde/Wonder-Griffin/TraXLMistralForCausalLM",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TraXLMistralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"dropout": 0.1,
|
| 7 |
+
"dynamic_routing": true,
|
| 8 |
+
"ff_expansion_factor": 4,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "LABEL_0",
|
| 12 |
+
"1": "LABEL_1",
|
| 13 |
+
"2": "LABEL_2",
|
| 14 |
+
"3": "LABEL_3",
|
| 15 |
+
"4": "LABEL_4"
|
| 16 |
+
},
|
| 17 |
+
"is_decoder": true,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"LABEL_0": 0,
|
| 20 |
+
"LABEL_1": 1,
|
| 21 |
+
"LABEL_2": 2,
|
| 22 |
+
"LABEL_3": 3,
|
| 23 |
+
"LABEL_4": 4
|
| 24 |
+
},
|
| 25 |
+
"max_computation_steps": 5,
|
| 26 |
+
"max_len": 256,
|
| 27 |
+
"memory_size": 256,
|
| 28 |
+
"model_type": "TraXLMistral",
|
| 29 |
+
"n_head": 4,
|
| 30 |
+
"n_layer": 4,
|
| 31 |
+
"rnn_units": 128,
|
| 32 |
+
"sparse_attention": true,
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.44.2",
|
| 35 |
+
"vocab_size": 50257
|
| 36 |
+
}
|
checkpoint-1148/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1148/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b6be909997b84a1a51472ad088079b2e6f7b8dc4ce6ed9f7a2426ab9719a1f5
|
| 3 |
+
size 194528200
|
checkpoint-1148/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e6fdcafd1487f22b973c6ad3139d46de43960c0411739503727ba24fb6f25dd
|
| 3 |
+
size 369911930
|
checkpoint-1148/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcf3aa11c7c8e9b65fbf8b69244e1780cf5f625d8d7ee3b83be7ab09e73bba15
|
| 3 |
+
size 14244
|
checkpoint-1148/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f321e90012081dffbd78dda3b4707d6d02928eed4f0464483697f03281263dc3
|
| 3 |
+
size 1064
|
checkpoint-1148/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
checkpoint-1148/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1148/tokenizer_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"bos_token": "<|endoftext|>",
|
| 14 |
+
"clean_up_tokenization_spaces": true,
|
| 15 |
+
"eos_token": "<|endoftext|>",
|
| 16 |
+
"model_max_length": 1024,
|
| 17 |
+
"pad_token": "<|endoftext|>",
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>"
|
| 20 |
+
}
|
checkpoint-1148/trainer_state.json
ADDED
|
@@ -0,0 +1,831 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1148,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008710801393728223,
|
| 13 |
+
"grad_norm": 96.23190307617188,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 10.808,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.017421602787456445,
|
| 20 |
+
"grad_norm": 463.5845642089844,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 10.8147,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02613240418118467,
|
| 27 |
+
"grad_norm": 0.10578063130378723,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 10.8077,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03484320557491289,
|
| 34 |
+
"grad_norm": 0.09780663996934891,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.8218,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04355400696864112,
|
| 41 |
+
"grad_norm": 1469.5677490234375,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.8079,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05226480836236934,
|
| 48 |
+
"grad_norm": 0.09817048907279968,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 10.8204,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06097560975609756,
|
| 55 |
+
"grad_norm": 0.09422990679740906,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 10.8159,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06968641114982578,
|
| 62 |
+
"grad_norm": 0.09545309096574783,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 10.811,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.078397212543554,
|
| 69 |
+
"grad_norm": 0.08887256681919098,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 10.8048,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08710801393728224,
|
| 76 |
+
"grad_norm": 0.0909484326839447,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 10.8126,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09581881533101046,
|
| 83 |
+
"grad_norm": 0.09697296470403671,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 10.8056,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10452961672473868,
|
| 90 |
+
"grad_norm": 0.09927204251289368,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 11.0541,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1132404181184669,
|
| 97 |
+
"grad_norm": 0.09392143785953522,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 10.8073,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.12195121951219512,
|
| 104 |
+
"grad_norm": 0.09451345354318619,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 10.8041,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.13066202090592335,
|
| 111 |
+
"grad_norm": 0.09947695583105087,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 10.8141,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13937282229965156,
|
| 118 |
+
"grad_norm": 0.09803537279367447,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 10.805,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1480836236933798,
|
| 125 |
+
"grad_norm": 0.09703430533409119,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 10.8193,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.156794425087108,
|
| 132 |
+
"grad_norm": 0.09272942692041397,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 10.8151,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16550522648083624,
|
| 139 |
+
"grad_norm": 0.0950935110449791,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 10.8051,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.17421602787456447,
|
| 146 |
+
"grad_norm": 0.10270467400550842,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 10.8043,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.18292682926829268,
|
| 153 |
+
"grad_norm": 0.09881923347711563,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 10.8044,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1916376306620209,
|
| 160 |
+
"grad_norm": 0.10348394513130188,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 10.8036,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.20034843205574912,
|
| 167 |
+
"grad_norm": 0.09483418613672256,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 10.8046,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20905923344947736,
|
| 174 |
+
"grad_norm": 0.09620082378387451,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 10.807,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21777003484320556,
|
| 181 |
+
"grad_norm": 0.09709980338811874,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 10.8058,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2264808362369338,
|
| 188 |
+
"grad_norm": 0.09373420476913452,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 10.8044,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.23519163763066203,
|
| 195 |
+
"grad_norm": 0.09700135141611099,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 10.8168,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.24390243902439024,
|
| 202 |
+
"grad_norm": 0.09450981020927429,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 10.8043,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.25261324041811845,
|
| 209 |
+
"grad_norm": 0.09983480721712112,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 10.8037,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2613240418118467,
|
| 216 |
+
"grad_norm": 0.09153986722230911,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 10.8041,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2700348432055749,
|
| 223 |
+
"grad_norm": 0.09664902836084366,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 10.805,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2787456445993031,
|
| 230 |
+
"grad_norm": 0.08636524528265,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 10.8049,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2874564459930314,
|
| 237 |
+
"grad_norm": 0.0945158526301384,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 10.8047,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2961672473867596,
|
| 244 |
+
"grad_norm": 0.09447965025901794,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 10.8044,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3048780487804878,
|
| 251 |
+
"grad_norm": 0.0949367955327034,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 10.8038,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.313588850174216,
|
| 258 |
+
"grad_norm": 0.10188119113445282,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 10.8037,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.32229965156794427,
|
| 265 |
+
"grad_norm": 0.09333271533250809,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 10.8029,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3310104529616725,
|
| 272 |
+
"grad_norm": 0.10046978294849396,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 10.8037,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3397212543554007,
|
| 279 |
+
"grad_norm": 0.09508766233921051,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 10.8065,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.34843205574912894,
|
| 286 |
+
"grad_norm": 0.09448560327291489,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 10.8058,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.35714285714285715,
|
| 293 |
+
"grad_norm": 0.09986421465873718,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 14.4299,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.36585365853658536,
|
| 300 |
+
"grad_norm": 0.10298220068216324,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 10.8035,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.37456445993031356,
|
| 307 |
+
"grad_norm": 0.09236953407526016,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 10.8309,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3832752613240418,
|
| 314 |
+
"grad_norm": 0.09000707417726517,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 10.8027,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.39198606271777003,
|
| 321 |
+
"grad_norm": 0.09554922580718994,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 10.8045,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.40069686411149824,
|
| 328 |
+
"grad_norm": 130.35623168945312,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 10.8034,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.4094076655052265,
|
| 335 |
+
"grad_norm": 0.09580319374799728,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 10.8022,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4181184668989547,
|
| 342 |
+
"grad_norm": 0.10328342765569687,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 10.8087,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4268292682926829,
|
| 349 |
+
"grad_norm": 0.09383571892976761,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 10.8027,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4355400696864111,
|
| 356 |
+
"grad_norm": 61532.49609375,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 15.4887,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4442508710801394,
|
| 363 |
+
"grad_norm": 0.09285067766904831,
|
| 364 |
+
"learning_rate": 9.966032608695653e-06,
|
| 365 |
+
"loss": 10.8028,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4529616724738676,
|
| 370 |
+
"grad_norm": 0.09329842031002045,
|
| 371 |
+
"learning_rate": 9.932065217391306e-06,
|
| 372 |
+
"loss": 10.8969,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4616724738675958,
|
| 377 |
+
"grad_norm": 0.09399975091218948,
|
| 378 |
+
"learning_rate": 9.898097826086957e-06,
|
| 379 |
+
"loss": 10.8297,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.47038327526132406,
|
| 384 |
+
"grad_norm": 1130.7960205078125,
|
| 385 |
+
"learning_rate": 9.86413043478261e-06,
|
| 386 |
+
"loss": 10.8358,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.47909407665505227,
|
| 391 |
+
"grad_norm": 0.0971442461013794,
|
| 392 |
+
"learning_rate": 9.830163043478262e-06,
|
| 393 |
+
"loss": 10.8641,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4878048780487805,
|
| 398 |
+
"grad_norm": 0.09969327598810196,
|
| 399 |
+
"learning_rate": 9.796195652173915e-06,
|
| 400 |
+
"loss": 10.8124,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4965156794425087,
|
| 405 |
+
"grad_norm": 0.0888582393527031,
|
| 406 |
+
"learning_rate": 9.762228260869566e-06,
|
| 407 |
+
"loss": 10.9576,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5052264808362369,
|
| 412 |
+
"grad_norm": 0.09847419708967209,
|
| 413 |
+
"learning_rate": 9.728260869565218e-06,
|
| 414 |
+
"loss": 10.8022,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5139372822299652,
|
| 419 |
+
"grad_norm": 0.0947929099202156,
|
| 420 |
+
"learning_rate": 9.694293478260869e-06,
|
| 421 |
+
"loss": 10.8011,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5226480836236934,
|
| 426 |
+
"grad_norm": 9.681638717651367,
|
| 427 |
+
"learning_rate": 9.660326086956523e-06,
|
| 428 |
+
"loss": 10.8015,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5313588850174216,
|
| 433 |
+
"grad_norm": 0.09560558199882507,
|
| 434 |
+
"learning_rate": 9.626358695652174e-06,
|
| 435 |
+
"loss": 13.2261,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5400696864111498,
|
| 440 |
+
"grad_norm": 0.09392021596431732,
|
| 441 |
+
"learning_rate": 9.592391304347827e-06,
|
| 442 |
+
"loss": 10.802,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5487804878048781,
|
| 447 |
+
"grad_norm": 0.10044202208518982,
|
| 448 |
+
"learning_rate": 9.558423913043478e-06,
|
| 449 |
+
"loss": 10.8005,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5574912891986062,
|
| 454 |
+
"grad_norm": 0.09005896002054214,
|
| 455 |
+
"learning_rate": 9.524456521739132e-06,
|
| 456 |
+
"loss": 10.802,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5662020905923345,
|
| 461 |
+
"grad_norm": 0.09686450660228729,
|
| 462 |
+
"learning_rate": 9.490489130434783e-06,
|
| 463 |
+
"loss": 10.801,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5749128919860628,
|
| 468 |
+
"grad_norm": 0.09321983903646469,
|
| 469 |
+
"learning_rate": 9.456521739130436e-06,
|
| 470 |
+
"loss": 10.8008,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5836236933797909,
|
| 475 |
+
"grad_norm": 0.0884283185005188,
|
| 476 |
+
"learning_rate": 9.422554347826087e-06,
|
| 477 |
+
"loss": 10.801,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5923344947735192,
|
| 482 |
+
"grad_norm": 0.09928812086582184,
|
| 483 |
+
"learning_rate": 9.388586956521741e-06,
|
| 484 |
+
"loss": 10.7998,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6010452961672473,
|
| 489 |
+
"grad_norm": 0.09434370696544647,
|
| 490 |
+
"learning_rate": 9.354619565217392e-06,
|
| 491 |
+
"loss": 10.8007,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6097560975609756,
|
| 496 |
+
"grad_norm": 0.09228064864873886,
|
| 497 |
+
"learning_rate": 9.320652173913044e-06,
|
| 498 |
+
"loss": 10.8002,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6184668989547039,
|
| 503 |
+
"grad_norm": 0.09412040561437607,
|
| 504 |
+
"learning_rate": 9.286684782608695e-06,
|
| 505 |
+
"loss": 10.8009,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.627177700348432,
|
| 510 |
+
"grad_norm": 0.0975504145026207,
|
| 511 |
+
"learning_rate": 9.25271739130435e-06,
|
| 512 |
+
"loss": 10.8002,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6358885017421603,
|
| 517 |
+
"grad_norm": 0.10332240164279938,
|
| 518 |
+
"learning_rate": 9.21875e-06,
|
| 519 |
+
"loss": 10.8001,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6445993031358885,
|
| 524 |
+
"grad_norm": 0.09318797290325165,
|
| 525 |
+
"learning_rate": 9.184782608695653e-06,
|
| 526 |
+
"loss": 10.8001,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6533101045296167,
|
| 531 |
+
"grad_norm": 0.09171754121780396,
|
| 532 |
+
"learning_rate": 9.150815217391306e-06,
|
| 533 |
+
"loss": 10.7999,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.662020905923345,
|
| 538 |
+
"grad_norm": 0.0921366959810257,
|
| 539 |
+
"learning_rate": 9.116847826086958e-06,
|
| 540 |
+
"loss": 10.8003,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6707317073170732,
|
| 545 |
+
"grad_norm": 0.09661433100700378,
|
| 546 |
+
"learning_rate": 9.08288043478261e-06,
|
| 547 |
+
"loss": 10.8001,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6794425087108014,
|
| 552 |
+
"grad_norm": 0.09323884546756744,
|
| 553 |
+
"learning_rate": 9.048913043478262e-06,
|
| 554 |
+
"loss": 10.7997,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6881533101045296,
|
| 559 |
+
"grad_norm": 0.08752495795488358,
|
| 560 |
+
"learning_rate": 9.014945652173914e-06,
|
| 561 |
+
"loss": 10.8,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6968641114982579,
|
| 566 |
+
"grad_norm": 0.09087631851434708,
|
| 567 |
+
"learning_rate": 8.980978260869567e-06,
|
| 568 |
+
"loss": 10.798,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.705574912891986,
|
| 573 |
+
"grad_norm": 0.09523475170135498,
|
| 574 |
+
"learning_rate": 8.947010869565218e-06,
|
| 575 |
+
"loss": 10.7994,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7142857142857143,
|
| 580 |
+
"grad_norm": 0.09918010234832764,
|
| 581 |
+
"learning_rate": 8.91304347826087e-06,
|
| 582 |
+
"loss": 10.7996,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.7229965156794426,
|
| 587 |
+
"grad_norm": 0.09539210051298141,
|
| 588 |
+
"learning_rate": 8.879076086956523e-06,
|
| 589 |
+
"loss": 10.7993,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.7317073170731707,
|
| 594 |
+
"grad_norm": 0.09296409040689468,
|
| 595 |
+
"learning_rate": 8.845108695652174e-06,
|
| 596 |
+
"loss": 10.7993,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.740418118466899,
|
| 601 |
+
"grad_norm": 0.08845611661672592,
|
| 602 |
+
"learning_rate": 8.811141304347827e-06,
|
| 603 |
+
"loss": 10.7993,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.7491289198606271,
|
| 608 |
+
"grad_norm": 0.0954870954155922,
|
| 609 |
+
"learning_rate": 8.77717391304348e-06,
|
| 610 |
+
"loss": 10.7981,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7578397212543554,
|
| 615 |
+
"grad_norm": 0.09371493011713028,
|
| 616 |
+
"learning_rate": 8.743206521739132e-06,
|
| 617 |
+
"loss": 10.7984,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7665505226480837,
|
| 622 |
+
"grad_norm": 0.09903618693351746,
|
| 623 |
+
"learning_rate": 8.709239130434783e-06,
|
| 624 |
+
"loss": 10.7987,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7752613240418118,
|
| 629 |
+
"grad_norm": 0.08991510421037674,
|
| 630 |
+
"learning_rate": 8.675271739130435e-06,
|
| 631 |
+
"loss": 10.7976,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7839721254355401,
|
| 636 |
+
"grad_norm": 0.09141765534877777,
|
| 637 |
+
"learning_rate": 8.641304347826088e-06,
|
| 638 |
+
"loss": 10.799,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7926829268292683,
|
| 643 |
+
"grad_norm": 0.09670023620128632,
|
| 644 |
+
"learning_rate": 8.60733695652174e-06,
|
| 645 |
+
"loss": 10.7973,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8013937282229965,
|
| 650 |
+
"grad_norm": 0.09337509423494339,
|
| 651 |
+
"learning_rate": 8.573369565217391e-06,
|
| 652 |
+
"loss": 10.7985,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.8101045296167247,
|
| 657 |
+
"grad_norm": 0.09618163853883743,
|
| 658 |
+
"learning_rate": 8.539402173913044e-06,
|
| 659 |
+
"loss": 10.7989,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.818815331010453,
|
| 664 |
+
"grad_norm": 0.09537822753190994,
|
| 665 |
+
"learning_rate": 8.505434782608697e-06,
|
| 666 |
+
"loss": 10.7981,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.8275261324041812,
|
| 671 |
+
"grad_norm": 0.09312313050031662,
|
| 672 |
+
"learning_rate": 8.47146739130435e-06,
|
| 673 |
+
"loss": 10.7984,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.8362369337979094,
|
| 678 |
+
"grad_norm": 0.09118187427520752,
|
| 679 |
+
"learning_rate": 8.4375e-06,
|
| 680 |
+
"loss": 10.7981,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.8449477351916377,
|
| 685 |
+
"grad_norm": 0.0919727087020874,
|
| 686 |
+
"learning_rate": 8.403532608695653e-06,
|
| 687 |
+
"loss": 10.7978,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8536585365853658,
|
| 692 |
+
"grad_norm": 0.09706488996744156,
|
| 693 |
+
"learning_rate": 8.369565217391305e-06,
|
| 694 |
+
"loss": 10.7979,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8623693379790941,
|
| 699 |
+
"grad_norm": 0.0944119319319725,
|
| 700 |
+
"learning_rate": 8.335597826086958e-06,
|
| 701 |
+
"loss": 10.7983,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8710801393728222,
|
| 706 |
+
"grad_norm": 0.1016387864947319,
|
| 707 |
+
"learning_rate": 8.301630434782609e-06,
|
| 708 |
+
"loss": 10.7973,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8797909407665505,
|
| 713 |
+
"grad_norm": 0.10198845714330673,
|
| 714 |
+
"learning_rate": 8.267663043478261e-06,
|
| 715 |
+
"loss": 10.7984,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8885017421602788,
|
| 720 |
+
"grad_norm": 0.10102390497922897,
|
| 721 |
+
"learning_rate": 8.233695652173914e-06,
|
| 722 |
+
"loss": 10.7971,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8972125435540069,
|
| 727 |
+
"grad_norm": 0.0972791463136673,
|
| 728 |
+
"learning_rate": 8.199728260869567e-06,
|
| 729 |
+
"loss": 10.7971,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9059233449477352,
|
| 734 |
+
"grad_norm": 0.09343789517879486,
|
| 735 |
+
"learning_rate": 8.165760869565218e-06,
|
| 736 |
+
"loss": 10.7978,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.9146341463414634,
|
| 741 |
+
"grad_norm": 0.09384151548147202,
|
| 742 |
+
"learning_rate": 8.13179347826087e-06,
|
| 743 |
+
"loss": 10.797,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.9233449477351916,
|
| 748 |
+
"grad_norm": 0.09324201196432114,
|
| 749 |
+
"learning_rate": 8.097826086956523e-06,
|
| 750 |
+
"loss": 10.7969,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.9320557491289199,
|
| 755 |
+
"grad_norm": 0.09341388940811157,
|
| 756 |
+
"learning_rate": 8.063858695652175e-06,
|
| 757 |
+
"loss": 10.7972,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9407665505226481,
|
| 762 |
+
"grad_norm": 0.09563016146421432,
|
| 763 |
+
"learning_rate": 8.029891304347826e-06,
|
| 764 |
+
"loss": 10.7978,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9494773519163763,
|
| 769 |
+
"grad_norm": 0.09752058237791061,
|
| 770 |
+
"learning_rate": 7.995923913043479e-06,
|
| 771 |
+
"loss": 10.7962,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9581881533101045,
|
| 776 |
+
"grad_norm": 0.09652635455131531,
|
| 777 |
+
"learning_rate": 7.961956521739131e-06,
|
| 778 |
+
"loss": 10.7968,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9668989547038328,
|
| 783 |
+
"grad_norm": 0.097730852663517,
|
| 784 |
+
"learning_rate": 7.927989130434784e-06,
|
| 785 |
+
"loss": 10.7967,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.975609756097561,
|
| 790 |
+
"grad_norm": 0.09178122133016586,
|
| 791 |
+
"learning_rate": 7.894021739130435e-06,
|
| 792 |
+
"loss": 10.7963,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9843205574912892,
|
| 797 |
+
"grad_norm": 0.09718208760023117,
|
| 798 |
+
"learning_rate": 7.860054347826088e-06,
|
| 799 |
+
"loss": 10.7967,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9930313588850174,
|
| 804 |
+
"grad_norm": 0.09228435158729553,
|
| 805 |
+
"learning_rate": 7.82608695652174e-06,
|
| 806 |
+
"loss": 10.7973,
|
| 807 |
+
"step": 1140
|
| 808 |
+
}
|
| 809 |
+
],
|
| 810 |
+
"logging_steps": 10,
|
| 811 |
+
"max_steps": 3444,
|
| 812 |
+
"num_input_tokens_seen": 0,
|
| 813 |
+
"num_train_epochs": 3,
|
| 814 |
+
"save_steps": 500,
|
| 815 |
+
"stateful_callbacks": {
|
| 816 |
+
"TrainerControl": {
|
| 817 |
+
"args": {
|
| 818 |
+
"should_epoch_stop": false,
|
| 819 |
+
"should_evaluate": false,
|
| 820 |
+
"should_log": false,
|
| 821 |
+
"should_save": true,
|
| 822 |
+
"should_training_stop": false
|
| 823 |
+
},
|
| 824 |
+
"attributes": {}
|
| 825 |
+
}
|
| 826 |
+
},
|
| 827 |
+
"total_flos": 565862839314432.0,
|
| 828 |
+
"train_batch_size": 32,
|
| 829 |
+
"trial_name": null,
|
| 830 |
+
"trial_params": null
|
| 831 |
+
}
|
checkpoint-1148/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90c67833faf31d4402ee0893264f8627505d9d95e6510f92d119f2d25dc10219
|
| 3 |
+
size 5176
|
checkpoint-1148/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2296/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "C:/Users/wonde/Wonder-Griffin/TraXLMistralForCausalLM",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TraXLMistralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"dropout": 0.1,
|
| 7 |
+
"dynamic_routing": true,
|
| 8 |
+
"ff_expansion_factor": 4,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "LABEL_0",
|
| 12 |
+
"1": "LABEL_1",
|
| 13 |
+
"2": "LABEL_2",
|
| 14 |
+
"3": "LABEL_3",
|
| 15 |
+
"4": "LABEL_4"
|
| 16 |
+
},
|
| 17 |
+
"is_decoder": true,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"LABEL_0": 0,
|
| 20 |
+
"LABEL_1": 1,
|
| 21 |
+
"LABEL_2": 2,
|
| 22 |
+
"LABEL_3": 3,
|
| 23 |
+
"LABEL_4": 4
|
| 24 |
+
},
|
| 25 |
+
"max_computation_steps": 5,
|
| 26 |
+
"max_len": 256,
|
| 27 |
+
"memory_size": 256,
|
| 28 |
+
"model_type": "TraXLMistral",
|
| 29 |
+
"n_head": 4,
|
| 30 |
+
"n_layer": 4,
|
| 31 |
+
"rnn_units": 128,
|
| 32 |
+
"sparse_attention": true,
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.44.2",
|
| 35 |
+
"vocab_size": 50257
|
| 36 |
+
}
|
checkpoint-2296/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2296/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fad2d4d6976258d643de42221c8bd88f7acc23dbd7f1978091ecb567ef1918f
|
| 3 |
+
size 194528200
|
checkpoint-2296/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab969e3bb92cda708fbff7bc54e939922d2975651349fa2f2fbb8c501f9b1160
|
| 3 |
+
size 369911930
|
checkpoint-2296/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c90750f83520c2b4a951eb3e729e4f3af09797e5ebb7d38846768f59d74a12ce
|
| 3 |
+
size 14244
|
checkpoint-2296/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22a75b57ea4043ff957c345c8e36e02e0cf3afa86f707b8644b3e062e6b30cf0
|
| 3 |
+
size 1064
|
checkpoint-2296/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
checkpoint-2296/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2296/tokenizer_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"bos_token": "<|endoftext|>",
|
| 14 |
+
"clean_up_tokenization_spaces": true,
|
| 15 |
+
"eos_token": "<|endoftext|>",
|
| 16 |
+
"model_max_length": 1024,
|
| 17 |
+
"pad_token": "<|endoftext|>",
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>"
|
| 20 |
+
}
|
checkpoint-2296/trainer_state.json
ADDED
|
@@ -0,0 +1,1636 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2296,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008710801393728223,
|
| 13 |
+
"grad_norm": 96.23190307617188,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 10.808,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.017421602787456445,
|
| 20 |
+
"grad_norm": 463.5845642089844,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 10.8147,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02613240418118467,
|
| 27 |
+
"grad_norm": 0.10578063130378723,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 10.8077,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03484320557491289,
|
| 34 |
+
"grad_norm": 0.09780663996934891,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.8218,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04355400696864112,
|
| 41 |
+
"grad_norm": 1469.5677490234375,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.8079,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05226480836236934,
|
| 48 |
+
"grad_norm": 0.09817048907279968,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 10.8204,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06097560975609756,
|
| 55 |
+
"grad_norm": 0.09422990679740906,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 10.8159,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06968641114982578,
|
| 62 |
+
"grad_norm": 0.09545309096574783,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 10.811,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.078397212543554,
|
| 69 |
+
"grad_norm": 0.08887256681919098,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 10.8048,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08710801393728224,
|
| 76 |
+
"grad_norm": 0.0909484326839447,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 10.8126,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09581881533101046,
|
| 83 |
+
"grad_norm": 0.09697296470403671,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 10.8056,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10452961672473868,
|
| 90 |
+
"grad_norm": 0.09927204251289368,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 11.0541,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1132404181184669,
|
| 97 |
+
"grad_norm": 0.09392143785953522,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 10.8073,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.12195121951219512,
|
| 104 |
+
"grad_norm": 0.09451345354318619,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 10.8041,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.13066202090592335,
|
| 111 |
+
"grad_norm": 0.09947695583105087,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 10.8141,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13937282229965156,
|
| 118 |
+
"grad_norm": 0.09803537279367447,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 10.805,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1480836236933798,
|
| 125 |
+
"grad_norm": 0.09703430533409119,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 10.8193,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.156794425087108,
|
| 132 |
+
"grad_norm": 0.09272942692041397,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 10.8151,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16550522648083624,
|
| 139 |
+
"grad_norm": 0.0950935110449791,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 10.8051,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.17421602787456447,
|
| 146 |
+
"grad_norm": 0.10270467400550842,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 10.8043,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.18292682926829268,
|
| 153 |
+
"grad_norm": 0.09881923347711563,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 10.8044,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1916376306620209,
|
| 160 |
+
"grad_norm": 0.10348394513130188,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 10.8036,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.20034843205574912,
|
| 167 |
+
"grad_norm": 0.09483418613672256,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 10.8046,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20905923344947736,
|
| 174 |
+
"grad_norm": 0.09620082378387451,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 10.807,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21777003484320556,
|
| 181 |
+
"grad_norm": 0.09709980338811874,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 10.8058,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2264808362369338,
|
| 188 |
+
"grad_norm": 0.09373420476913452,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 10.8044,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.23519163763066203,
|
| 195 |
+
"grad_norm": 0.09700135141611099,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 10.8168,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.24390243902439024,
|
| 202 |
+
"grad_norm": 0.09450981020927429,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 10.8043,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.25261324041811845,
|
| 209 |
+
"grad_norm": 0.09983480721712112,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 10.8037,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2613240418118467,
|
| 216 |
+
"grad_norm": 0.09153986722230911,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 10.8041,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2700348432055749,
|
| 223 |
+
"grad_norm": 0.09664902836084366,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 10.805,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2787456445993031,
|
| 230 |
+
"grad_norm": 0.08636524528265,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 10.8049,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2874564459930314,
|
| 237 |
+
"grad_norm": 0.0945158526301384,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 10.8047,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2961672473867596,
|
| 244 |
+
"grad_norm": 0.09447965025901794,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 10.8044,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3048780487804878,
|
| 251 |
+
"grad_norm": 0.0949367955327034,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 10.8038,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.313588850174216,
|
| 258 |
+
"grad_norm": 0.10188119113445282,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 10.8037,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.32229965156794427,
|
| 265 |
+
"grad_norm": 0.09333271533250809,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 10.8029,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3310104529616725,
|
| 272 |
+
"grad_norm": 0.10046978294849396,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 10.8037,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3397212543554007,
|
| 279 |
+
"grad_norm": 0.09508766233921051,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 10.8065,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.34843205574912894,
|
| 286 |
+
"grad_norm": 0.09448560327291489,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 10.8058,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.35714285714285715,
|
| 293 |
+
"grad_norm": 0.09986421465873718,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 14.4299,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.36585365853658536,
|
| 300 |
+
"grad_norm": 0.10298220068216324,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 10.8035,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.37456445993031356,
|
| 307 |
+
"grad_norm": 0.09236953407526016,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 10.8309,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3832752613240418,
|
| 314 |
+
"grad_norm": 0.09000707417726517,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 10.8027,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.39198606271777003,
|
| 321 |
+
"grad_norm": 0.09554922580718994,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 10.8045,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.40069686411149824,
|
| 328 |
+
"grad_norm": 130.35623168945312,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 10.8034,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.4094076655052265,
|
| 335 |
+
"grad_norm": 0.09580319374799728,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 10.8022,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4181184668989547,
|
| 342 |
+
"grad_norm": 0.10328342765569687,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 10.8087,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4268292682926829,
|
| 349 |
+
"grad_norm": 0.09383571892976761,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 10.8027,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4355400696864111,
|
| 356 |
+
"grad_norm": 61532.49609375,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 15.4887,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4442508710801394,
|
| 363 |
+
"grad_norm": 0.09285067766904831,
|
| 364 |
+
"learning_rate": 9.966032608695653e-06,
|
| 365 |
+
"loss": 10.8028,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4529616724738676,
|
| 370 |
+
"grad_norm": 0.09329842031002045,
|
| 371 |
+
"learning_rate": 9.932065217391306e-06,
|
| 372 |
+
"loss": 10.8969,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4616724738675958,
|
| 377 |
+
"grad_norm": 0.09399975091218948,
|
| 378 |
+
"learning_rate": 9.898097826086957e-06,
|
| 379 |
+
"loss": 10.8297,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.47038327526132406,
|
| 384 |
+
"grad_norm": 1130.7960205078125,
|
| 385 |
+
"learning_rate": 9.86413043478261e-06,
|
| 386 |
+
"loss": 10.8358,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.47909407665505227,
|
| 391 |
+
"grad_norm": 0.0971442461013794,
|
| 392 |
+
"learning_rate": 9.830163043478262e-06,
|
| 393 |
+
"loss": 10.8641,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4878048780487805,
|
| 398 |
+
"grad_norm": 0.09969327598810196,
|
| 399 |
+
"learning_rate": 9.796195652173915e-06,
|
| 400 |
+
"loss": 10.8124,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4965156794425087,
|
| 405 |
+
"grad_norm": 0.0888582393527031,
|
| 406 |
+
"learning_rate": 9.762228260869566e-06,
|
| 407 |
+
"loss": 10.9576,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5052264808362369,
|
| 412 |
+
"grad_norm": 0.09847419708967209,
|
| 413 |
+
"learning_rate": 9.728260869565218e-06,
|
| 414 |
+
"loss": 10.8022,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5139372822299652,
|
| 419 |
+
"grad_norm": 0.0947929099202156,
|
| 420 |
+
"learning_rate": 9.694293478260869e-06,
|
| 421 |
+
"loss": 10.8011,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5226480836236934,
|
| 426 |
+
"grad_norm": 9.681638717651367,
|
| 427 |
+
"learning_rate": 9.660326086956523e-06,
|
| 428 |
+
"loss": 10.8015,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5313588850174216,
|
| 433 |
+
"grad_norm": 0.09560558199882507,
|
| 434 |
+
"learning_rate": 9.626358695652174e-06,
|
| 435 |
+
"loss": 13.2261,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5400696864111498,
|
| 440 |
+
"grad_norm": 0.09392021596431732,
|
| 441 |
+
"learning_rate": 9.592391304347827e-06,
|
| 442 |
+
"loss": 10.802,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5487804878048781,
|
| 447 |
+
"grad_norm": 0.10044202208518982,
|
| 448 |
+
"learning_rate": 9.558423913043478e-06,
|
| 449 |
+
"loss": 10.8005,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5574912891986062,
|
| 454 |
+
"grad_norm": 0.09005896002054214,
|
| 455 |
+
"learning_rate": 9.524456521739132e-06,
|
| 456 |
+
"loss": 10.802,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5662020905923345,
|
| 461 |
+
"grad_norm": 0.09686450660228729,
|
| 462 |
+
"learning_rate": 9.490489130434783e-06,
|
| 463 |
+
"loss": 10.801,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5749128919860628,
|
| 468 |
+
"grad_norm": 0.09321983903646469,
|
| 469 |
+
"learning_rate": 9.456521739130436e-06,
|
| 470 |
+
"loss": 10.8008,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5836236933797909,
|
| 475 |
+
"grad_norm": 0.0884283185005188,
|
| 476 |
+
"learning_rate": 9.422554347826087e-06,
|
| 477 |
+
"loss": 10.801,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5923344947735192,
|
| 482 |
+
"grad_norm": 0.09928812086582184,
|
| 483 |
+
"learning_rate": 9.388586956521741e-06,
|
| 484 |
+
"loss": 10.7998,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6010452961672473,
|
| 489 |
+
"grad_norm": 0.09434370696544647,
|
| 490 |
+
"learning_rate": 9.354619565217392e-06,
|
| 491 |
+
"loss": 10.8007,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6097560975609756,
|
| 496 |
+
"grad_norm": 0.09228064864873886,
|
| 497 |
+
"learning_rate": 9.320652173913044e-06,
|
| 498 |
+
"loss": 10.8002,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6184668989547039,
|
| 503 |
+
"grad_norm": 0.09412040561437607,
|
| 504 |
+
"learning_rate": 9.286684782608695e-06,
|
| 505 |
+
"loss": 10.8009,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.627177700348432,
|
| 510 |
+
"grad_norm": 0.0975504145026207,
|
| 511 |
+
"learning_rate": 9.25271739130435e-06,
|
| 512 |
+
"loss": 10.8002,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6358885017421603,
|
| 517 |
+
"grad_norm": 0.10332240164279938,
|
| 518 |
+
"learning_rate": 9.21875e-06,
|
| 519 |
+
"loss": 10.8001,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6445993031358885,
|
| 524 |
+
"grad_norm": 0.09318797290325165,
|
| 525 |
+
"learning_rate": 9.184782608695653e-06,
|
| 526 |
+
"loss": 10.8001,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6533101045296167,
|
| 531 |
+
"grad_norm": 0.09171754121780396,
|
| 532 |
+
"learning_rate": 9.150815217391306e-06,
|
| 533 |
+
"loss": 10.7999,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.662020905923345,
|
| 538 |
+
"grad_norm": 0.0921366959810257,
|
| 539 |
+
"learning_rate": 9.116847826086958e-06,
|
| 540 |
+
"loss": 10.8003,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6707317073170732,
|
| 545 |
+
"grad_norm": 0.09661433100700378,
|
| 546 |
+
"learning_rate": 9.08288043478261e-06,
|
| 547 |
+
"loss": 10.8001,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6794425087108014,
|
| 552 |
+
"grad_norm": 0.09323884546756744,
|
| 553 |
+
"learning_rate": 9.048913043478262e-06,
|
| 554 |
+
"loss": 10.7997,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6881533101045296,
|
| 559 |
+
"grad_norm": 0.08752495795488358,
|
| 560 |
+
"learning_rate": 9.014945652173914e-06,
|
| 561 |
+
"loss": 10.8,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6968641114982579,
|
| 566 |
+
"grad_norm": 0.09087631851434708,
|
| 567 |
+
"learning_rate": 8.980978260869567e-06,
|
| 568 |
+
"loss": 10.798,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.705574912891986,
|
| 573 |
+
"grad_norm": 0.09523475170135498,
|
| 574 |
+
"learning_rate": 8.947010869565218e-06,
|
| 575 |
+
"loss": 10.7994,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7142857142857143,
|
| 580 |
+
"grad_norm": 0.09918010234832764,
|
| 581 |
+
"learning_rate": 8.91304347826087e-06,
|
| 582 |
+
"loss": 10.7996,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.7229965156794426,
|
| 587 |
+
"grad_norm": 0.09539210051298141,
|
| 588 |
+
"learning_rate": 8.879076086956523e-06,
|
| 589 |
+
"loss": 10.7993,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.7317073170731707,
|
| 594 |
+
"grad_norm": 0.09296409040689468,
|
| 595 |
+
"learning_rate": 8.845108695652174e-06,
|
| 596 |
+
"loss": 10.7993,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.740418118466899,
|
| 601 |
+
"grad_norm": 0.08845611661672592,
|
| 602 |
+
"learning_rate": 8.811141304347827e-06,
|
| 603 |
+
"loss": 10.7993,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.7491289198606271,
|
| 608 |
+
"grad_norm": 0.0954870954155922,
|
| 609 |
+
"learning_rate": 8.77717391304348e-06,
|
| 610 |
+
"loss": 10.7981,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7578397212543554,
|
| 615 |
+
"grad_norm": 0.09371493011713028,
|
| 616 |
+
"learning_rate": 8.743206521739132e-06,
|
| 617 |
+
"loss": 10.7984,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7665505226480837,
|
| 622 |
+
"grad_norm": 0.09903618693351746,
|
| 623 |
+
"learning_rate": 8.709239130434783e-06,
|
| 624 |
+
"loss": 10.7987,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7752613240418118,
|
| 629 |
+
"grad_norm": 0.08991510421037674,
|
| 630 |
+
"learning_rate": 8.675271739130435e-06,
|
| 631 |
+
"loss": 10.7976,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7839721254355401,
|
| 636 |
+
"grad_norm": 0.09141765534877777,
|
| 637 |
+
"learning_rate": 8.641304347826088e-06,
|
| 638 |
+
"loss": 10.799,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7926829268292683,
|
| 643 |
+
"grad_norm": 0.09670023620128632,
|
| 644 |
+
"learning_rate": 8.60733695652174e-06,
|
| 645 |
+
"loss": 10.7973,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8013937282229965,
|
| 650 |
+
"grad_norm": 0.09337509423494339,
|
| 651 |
+
"learning_rate": 8.573369565217391e-06,
|
| 652 |
+
"loss": 10.7985,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.8101045296167247,
|
| 657 |
+
"grad_norm": 0.09618163853883743,
|
| 658 |
+
"learning_rate": 8.539402173913044e-06,
|
| 659 |
+
"loss": 10.7989,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.818815331010453,
|
| 664 |
+
"grad_norm": 0.09537822753190994,
|
| 665 |
+
"learning_rate": 8.505434782608697e-06,
|
| 666 |
+
"loss": 10.7981,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.8275261324041812,
|
| 671 |
+
"grad_norm": 0.09312313050031662,
|
| 672 |
+
"learning_rate": 8.47146739130435e-06,
|
| 673 |
+
"loss": 10.7984,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.8362369337979094,
|
| 678 |
+
"grad_norm": 0.09118187427520752,
|
| 679 |
+
"learning_rate": 8.4375e-06,
|
| 680 |
+
"loss": 10.7981,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.8449477351916377,
|
| 685 |
+
"grad_norm": 0.0919727087020874,
|
| 686 |
+
"learning_rate": 8.403532608695653e-06,
|
| 687 |
+
"loss": 10.7978,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8536585365853658,
|
| 692 |
+
"grad_norm": 0.09706488996744156,
|
| 693 |
+
"learning_rate": 8.369565217391305e-06,
|
| 694 |
+
"loss": 10.7979,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8623693379790941,
|
| 699 |
+
"grad_norm": 0.0944119319319725,
|
| 700 |
+
"learning_rate": 8.335597826086958e-06,
|
| 701 |
+
"loss": 10.7983,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8710801393728222,
|
| 706 |
+
"grad_norm": 0.1016387864947319,
|
| 707 |
+
"learning_rate": 8.301630434782609e-06,
|
| 708 |
+
"loss": 10.7973,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8797909407665505,
|
| 713 |
+
"grad_norm": 0.10198845714330673,
|
| 714 |
+
"learning_rate": 8.267663043478261e-06,
|
| 715 |
+
"loss": 10.7984,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8885017421602788,
|
| 720 |
+
"grad_norm": 0.10102390497922897,
|
| 721 |
+
"learning_rate": 8.233695652173914e-06,
|
| 722 |
+
"loss": 10.7971,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8972125435540069,
|
| 727 |
+
"grad_norm": 0.0972791463136673,
|
| 728 |
+
"learning_rate": 8.199728260869567e-06,
|
| 729 |
+
"loss": 10.7971,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9059233449477352,
|
| 734 |
+
"grad_norm": 0.09343789517879486,
|
| 735 |
+
"learning_rate": 8.165760869565218e-06,
|
| 736 |
+
"loss": 10.7978,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.9146341463414634,
|
| 741 |
+
"grad_norm": 0.09384151548147202,
|
| 742 |
+
"learning_rate": 8.13179347826087e-06,
|
| 743 |
+
"loss": 10.797,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.9233449477351916,
|
| 748 |
+
"grad_norm": 0.09324201196432114,
|
| 749 |
+
"learning_rate": 8.097826086956523e-06,
|
| 750 |
+
"loss": 10.7969,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.9320557491289199,
|
| 755 |
+
"grad_norm": 0.09341388940811157,
|
| 756 |
+
"learning_rate": 8.063858695652175e-06,
|
| 757 |
+
"loss": 10.7972,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9407665505226481,
|
| 762 |
+
"grad_norm": 0.09563016146421432,
|
| 763 |
+
"learning_rate": 8.029891304347826e-06,
|
| 764 |
+
"loss": 10.7978,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9494773519163763,
|
| 769 |
+
"grad_norm": 0.09752058237791061,
|
| 770 |
+
"learning_rate": 7.995923913043479e-06,
|
| 771 |
+
"loss": 10.7962,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9581881533101045,
|
| 776 |
+
"grad_norm": 0.09652635455131531,
|
| 777 |
+
"learning_rate": 7.961956521739131e-06,
|
| 778 |
+
"loss": 10.7968,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9668989547038328,
|
| 783 |
+
"grad_norm": 0.097730852663517,
|
| 784 |
+
"learning_rate": 7.927989130434784e-06,
|
| 785 |
+
"loss": 10.7967,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.975609756097561,
|
| 790 |
+
"grad_norm": 0.09178122133016586,
|
| 791 |
+
"learning_rate": 7.894021739130435e-06,
|
| 792 |
+
"loss": 10.7963,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9843205574912892,
|
| 797 |
+
"grad_norm": 0.09718208760023117,
|
| 798 |
+
"learning_rate": 7.860054347826088e-06,
|
| 799 |
+
"loss": 10.7967,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9930313588850174,
|
| 804 |
+
"grad_norm": 0.09228435158729553,
|
| 805 |
+
"learning_rate": 7.82608695652174e-06,
|
| 806 |
+
"loss": 10.7973,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.0017421602787457,
|
| 811 |
+
"grad_norm": 0.08885621279478073,
|
| 812 |
+
"learning_rate": 7.792119565217393e-06,
|
| 813 |
+
"loss": 10.796,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.0104529616724738,
|
| 818 |
+
"grad_norm": 0.09165474027395248,
|
| 819 |
+
"learning_rate": 7.758152173913044e-06,
|
| 820 |
+
"loss": 10.7957,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.019163763066202,
|
| 825 |
+
"grad_norm": 0.09650120139122009,
|
| 826 |
+
"learning_rate": 7.724184782608696e-06,
|
| 827 |
+
"loss": 10.7967,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.0278745644599303,
|
| 832 |
+
"grad_norm": 0.10563351958990097,
|
| 833 |
+
"learning_rate": 7.690217391304349e-06,
|
| 834 |
+
"loss": 10.7963,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.0365853658536586,
|
| 839 |
+
"grad_norm": 0.09407645463943481,
|
| 840 |
+
"learning_rate": 7.656250000000001e-06,
|
| 841 |
+
"loss": 10.7958,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.0452961672473868,
|
| 846 |
+
"grad_norm": 0.100870780646801,
|
| 847 |
+
"learning_rate": 7.622282608695652e-06,
|
| 848 |
+
"loss": 10.7961,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.054006968641115,
|
| 853 |
+
"grad_norm": 0.09036817401647568,
|
| 854 |
+
"learning_rate": 7.588315217391306e-06,
|
| 855 |
+
"loss": 10.7965,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.0627177700348431,
|
| 860 |
+
"grad_norm": 0.09792335331439972,
|
| 861 |
+
"learning_rate": 7.5543478260869576e-06,
|
| 862 |
+
"loss": 10.7951,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.0714285714285714,
|
| 867 |
+
"grad_norm": 0.10280462354421616,
|
| 868 |
+
"learning_rate": 7.520380434782609e-06,
|
| 869 |
+
"loss": 10.796,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0801393728222997,
|
| 874 |
+
"grad_norm": 0.09284704178571701,
|
| 875 |
+
"learning_rate": 7.486413043478261e-06,
|
| 876 |
+
"loss": 10.796,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.088850174216028,
|
| 881 |
+
"grad_norm": 0.09495861083269119,
|
| 882 |
+
"learning_rate": 7.4524456521739145e-06,
|
| 883 |
+
"loss": 10.7954,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.0975609756097562,
|
| 888 |
+
"grad_norm": 0.0963548794388771,
|
| 889 |
+
"learning_rate": 7.418478260869566e-06,
|
| 890 |
+
"loss": 10.7954,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.1062717770034842,
|
| 895 |
+
"grad_norm": 0.10302776098251343,
|
| 896 |
+
"learning_rate": 7.384510869565218e-06,
|
| 897 |
+
"loss": 10.7946,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.1149825783972125,
|
| 902 |
+
"grad_norm": 0.09135697036981583,
|
| 903 |
+
"learning_rate": 7.35054347826087e-06,
|
| 904 |
+
"loss": 10.7951,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.1236933797909407,
|
| 909 |
+
"grad_norm": 0.09077905118465424,
|
| 910 |
+
"learning_rate": 7.316576086956522e-06,
|
| 911 |
+
"loss": 10.7949,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.132404181184669,
|
| 916 |
+
"grad_norm": 0.09623997658491135,
|
| 917 |
+
"learning_rate": 7.282608695652175e-06,
|
| 918 |
+
"loss": 10.795,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.1411149825783973,
|
| 923 |
+
"grad_norm": 0.09430749714374542,
|
| 924 |
+
"learning_rate": 7.248641304347827e-06,
|
| 925 |
+
"loss": 10.7954,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.1498257839721253,
|
| 930 |
+
"grad_norm": 0.09416501224040985,
|
| 931 |
+
"learning_rate": 7.2146739130434785e-06,
|
| 932 |
+
"loss": 10.795,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.1585365853658536,
|
| 937 |
+
"grad_norm": 0.09839834272861481,
|
| 938 |
+
"learning_rate": 7.180706521739131e-06,
|
| 939 |
+
"loss": 10.795,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.1672473867595818,
|
| 944 |
+
"grad_norm": 0.09414810687303543,
|
| 945 |
+
"learning_rate": 7.146739130434784e-06,
|
| 946 |
+
"loss": 10.7949,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.17595818815331,
|
| 951 |
+
"grad_norm": 0.09453831613063812,
|
| 952 |
+
"learning_rate": 7.1127717391304354e-06,
|
| 953 |
+
"loss": 10.942,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.1846689895470384,
|
| 958 |
+
"grad_norm": 637.3465576171875,
|
| 959 |
+
"learning_rate": 7.078804347826087e-06,
|
| 960 |
+
"loss": 10.8174,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.1933797909407666,
|
| 965 |
+
"grad_norm": 0.09487204998731613,
|
| 966 |
+
"learning_rate": 7.04483695652174e-06,
|
| 967 |
+
"loss": 16.214,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.202090592334495,
|
| 972 |
+
"grad_norm": 0.08683289587497711,
|
| 973 |
+
"learning_rate": 7.0108695652173915e-06,
|
| 974 |
+
"loss": 10.7961,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.210801393728223,
|
| 979 |
+
"grad_norm": 0.09138838201761246,
|
| 980 |
+
"learning_rate": 6.976902173913044e-06,
|
| 981 |
+
"loss": 10.7954,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.2195121951219512,
|
| 986 |
+
"grad_norm": 0.09047773480415344,
|
| 987 |
+
"learning_rate": 6.942934782608696e-06,
|
| 988 |
+
"loss": 10.7941,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.2282229965156795,
|
| 993 |
+
"grad_norm": 0.10373370349407196,
|
| 994 |
+
"learning_rate": 6.9089673913043485e-06,
|
| 995 |
+
"loss": 10.7933,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.2369337979094077,
|
| 1000 |
+
"grad_norm": 0.0910143107175827,
|
| 1001 |
+
"learning_rate": 6.875e-06,
|
| 1002 |
+
"loss": 10.7952,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.245644599303136,
|
| 1007 |
+
"grad_norm": 0.08921236544847488,
|
| 1008 |
+
"learning_rate": 6.841032608695653e-06,
|
| 1009 |
+
"loss": 10.7945,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.254355400696864,
|
| 1014 |
+
"grad_norm": 0.09373684972524643,
|
| 1015 |
+
"learning_rate": 6.8070652173913054e-06,
|
| 1016 |
+
"loss": 10.795,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.2630662020905923,
|
| 1021 |
+
"grad_norm": 0.09562421590089798,
|
| 1022 |
+
"learning_rate": 6.773097826086957e-06,
|
| 1023 |
+
"loss": 10.7948,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.2717770034843205,
|
| 1028 |
+
"grad_norm": 0.0909007340669632,
|
| 1029 |
+
"learning_rate": 6.739130434782609e-06,
|
| 1030 |
+
"loss": 10.7945,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.2804878048780488,
|
| 1035 |
+
"grad_norm": 0.09700188040733337,
|
| 1036 |
+
"learning_rate": 6.705163043478261e-06,
|
| 1037 |
+
"loss": 10.7947,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.289198606271777,
|
| 1042 |
+
"grad_norm": 0.09472058713436127,
|
| 1043 |
+
"learning_rate": 6.671195652173914e-06,
|
| 1044 |
+
"loss": 10.7942,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.297909407665505,
|
| 1049 |
+
"grad_norm": 0.09371300041675568,
|
| 1050 |
+
"learning_rate": 6.637228260869566e-06,
|
| 1051 |
+
"loss": 10.793,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.3066202090592334,
|
| 1056 |
+
"grad_norm": 0.08980853110551834,
|
| 1057 |
+
"learning_rate": 6.603260869565218e-06,
|
| 1058 |
+
"loss": 10.7936,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.3153310104529616,
|
| 1063 |
+
"grad_norm": 0.09769923985004425,
|
| 1064 |
+
"learning_rate": 6.569293478260869e-06,
|
| 1065 |
+
"loss": 10.7939,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.32404181184669,
|
| 1070 |
+
"grad_norm": 0.0956735908985138,
|
| 1071 |
+
"learning_rate": 6.535326086956523e-06,
|
| 1072 |
+
"loss": 10.7934,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.3327526132404182,
|
| 1077 |
+
"grad_norm": 0.09061311185359955,
|
| 1078 |
+
"learning_rate": 6.501358695652175e-06,
|
| 1079 |
+
"loss": 10.7937,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.3414634146341464,
|
| 1084 |
+
"grad_norm": 72412.171875,
|
| 1085 |
+
"learning_rate": 6.467391304347826e-06,
|
| 1086 |
+
"loss": 13.1839,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.3501742160278747,
|
| 1091 |
+
"grad_norm": 0.09127330780029297,
|
| 1092 |
+
"learning_rate": 6.433423913043478e-06,
|
| 1093 |
+
"loss": 13.2599,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.3588850174216027,
|
| 1098 |
+
"grad_norm": 0.09337225556373596,
|
| 1099 |
+
"learning_rate": 6.3994565217391316e-06,
|
| 1100 |
+
"loss": 10.8129,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.367595818815331,
|
| 1105 |
+
"grad_norm": 0.10567159950733185,
|
| 1106 |
+
"learning_rate": 6.365489130434783e-06,
|
| 1107 |
+
"loss": 10.7932,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3763066202090593,
|
| 1112 |
+
"grad_norm": 0.09582757949829102,
|
| 1113 |
+
"learning_rate": 6.331521739130435e-06,
|
| 1114 |
+
"loss": 10.7934,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.3850174216027875,
|
| 1119 |
+
"grad_norm": 0.09011757373809814,
|
| 1120 |
+
"learning_rate": 6.297554347826087e-06,
|
| 1121 |
+
"loss": 10.7934,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.3937282229965158,
|
| 1126 |
+
"grad_norm": 0.09461648017168045,
|
| 1127 |
+
"learning_rate": 6.26358695652174e-06,
|
| 1128 |
+
"loss": 10.7922,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.4024390243902438,
|
| 1133 |
+
"grad_norm": 0.0957835465669632,
|
| 1134 |
+
"learning_rate": 6.229619565217392e-06,
|
| 1135 |
+
"loss": 10.793,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.411149825783972,
|
| 1140 |
+
"grad_norm": 0.09626618772745132,
|
| 1141 |
+
"learning_rate": 6.195652173913044e-06,
|
| 1142 |
+
"loss": 10.7918,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.4198606271777003,
|
| 1147 |
+
"grad_norm": 0.09974712133407593,
|
| 1148 |
+
"learning_rate": 6.1616847826086955e-06,
|
| 1149 |
+
"loss": 10.7932,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.4285714285714286,
|
| 1154 |
+
"grad_norm": 0.09606435894966125,
|
| 1155 |
+
"learning_rate": 6.127717391304349e-06,
|
| 1156 |
+
"loss": 10.7926,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.4372822299651569,
|
| 1161 |
+
"grad_norm": 0.10175999999046326,
|
| 1162 |
+
"learning_rate": 6.093750000000001e-06,
|
| 1163 |
+
"loss": 10.7929,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.445993031358885,
|
| 1168 |
+
"grad_norm": 0.09452734142541885,
|
| 1169 |
+
"learning_rate": 6.0597826086956525e-06,
|
| 1170 |
+
"loss": 10.7936,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.4547038327526132,
|
| 1175 |
+
"grad_norm": 0.09544497728347778,
|
| 1176 |
+
"learning_rate": 6.025815217391305e-06,
|
| 1177 |
+
"loss": 10.793,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.4634146341463414,
|
| 1182 |
+
"grad_norm": 0.0867006853222847,
|
| 1183 |
+
"learning_rate": 5.991847826086957e-06,
|
| 1184 |
+
"loss": 10.7931,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.4721254355400697,
|
| 1189 |
+
"grad_norm": 0.09889491647481918,
|
| 1190 |
+
"learning_rate": 5.9578804347826094e-06,
|
| 1191 |
+
"loss": 10.7926,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.480836236933798,
|
| 1196 |
+
"grad_norm": 0.09840565174818039,
|
| 1197 |
+
"learning_rate": 5.923913043478261e-06,
|
| 1198 |
+
"loss": 10.7926,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.489547038327526,
|
| 1203 |
+
"grad_norm": 0.09129170328378677,
|
| 1204 |
+
"learning_rate": 5.889945652173914e-06,
|
| 1205 |
+
"loss": 10.7921,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.4982578397212545,
|
| 1210 |
+
"grad_norm": 0.09560216963291168,
|
| 1211 |
+
"learning_rate": 5.8559782608695656e-06,
|
| 1212 |
+
"loss": 10.7924,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.5069686411149825,
|
| 1217 |
+
"grad_norm": 0.09289251267910004,
|
| 1218 |
+
"learning_rate": 5.822010869565218e-06,
|
| 1219 |
+
"loss": 10.792,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.5156794425087108,
|
| 1224 |
+
"grad_norm": 0.09835352748632431,
|
| 1225 |
+
"learning_rate": 5.78804347826087e-06,
|
| 1226 |
+
"loss": 10.7917,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.524390243902439,
|
| 1231 |
+
"grad_norm": 0.10089673846960068,
|
| 1232 |
+
"learning_rate": 5.7540760869565225e-06,
|
| 1233 |
+
"loss": 10.7921,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.533101045296167,
|
| 1238 |
+
"grad_norm": 0.0979476273059845,
|
| 1239 |
+
"learning_rate": 5.720108695652174e-06,
|
| 1240 |
+
"loss": 10.7933,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.5418118466898956,
|
| 1245 |
+
"grad_norm": 0.10268489271402359,
|
| 1246 |
+
"learning_rate": 5.686141304347826e-06,
|
| 1247 |
+
"loss": 10.7946,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.5505226480836236,
|
| 1252 |
+
"grad_norm": 0.09039770066738129,
|
| 1253 |
+
"learning_rate": 5.652173913043479e-06,
|
| 1254 |
+
"loss": 10.7923,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.5592334494773519,
|
| 1259 |
+
"grad_norm": 0.09356367588043213,
|
| 1260 |
+
"learning_rate": 5.618206521739131e-06,
|
| 1261 |
+
"loss": 10.7918,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.5679442508710801,
|
| 1266 |
+
"grad_norm": 0.09169076383113861,
|
| 1267 |
+
"learning_rate": 5.584239130434783e-06,
|
| 1268 |
+
"loss": 10.7919,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.5766550522648084,
|
| 1273 |
+
"grad_norm": 0.09374472498893738,
|
| 1274 |
+
"learning_rate": 5.550271739130435e-06,
|
| 1275 |
+
"loss": 10.7922,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.5853658536585367,
|
| 1280 |
+
"grad_norm": 0.09261377900838852,
|
| 1281 |
+
"learning_rate": 5.516304347826087e-06,
|
| 1282 |
+
"loss": 10.7911,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.5940766550522647,
|
| 1287 |
+
"grad_norm": 0.0917358323931694,
|
| 1288 |
+
"learning_rate": 5.48233695652174e-06,
|
| 1289 |
+
"loss": 10.7926,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.6027874564459932,
|
| 1294 |
+
"grad_norm": 0.09259490668773651,
|
| 1295 |
+
"learning_rate": 5.448369565217392e-06,
|
| 1296 |
+
"loss": 10.7913,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.6114982578397212,
|
| 1301 |
+
"grad_norm": 0.09436757117509842,
|
| 1302 |
+
"learning_rate": 5.4144021739130434e-06,
|
| 1303 |
+
"loss": 10.7907,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.6202090592334495,
|
| 1308 |
+
"grad_norm": 0.0956592783331871,
|
| 1309 |
+
"learning_rate": 5.380434782608695e-06,
|
| 1310 |
+
"loss": 10.7925,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.6289198606271778,
|
| 1315 |
+
"grad_norm": 0.09467855840921402,
|
| 1316 |
+
"learning_rate": 5.346467391304349e-06,
|
| 1317 |
+
"loss": 10.7921,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.6376306620209058,
|
| 1322 |
+
"grad_norm": 0.0989622175693512,
|
| 1323 |
+
"learning_rate": 5.3125e-06,
|
| 1324 |
+
"loss": 10.7918,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.6463414634146343,
|
| 1329 |
+
"grad_norm": 0.09564294666051865,
|
| 1330 |
+
"learning_rate": 5.278532608695652e-06,
|
| 1331 |
+
"loss": 10.7913,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.6550522648083623,
|
| 1336 |
+
"grad_norm": 0.09419780969619751,
|
| 1337 |
+
"learning_rate": 5.244565217391306e-06,
|
| 1338 |
+
"loss": 10.7917,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.6637630662020906,
|
| 1343 |
+
"grad_norm": 0.0949205756187439,
|
| 1344 |
+
"learning_rate": 5.210597826086957e-06,
|
| 1345 |
+
"loss": 10.791,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.6724738675958188,
|
| 1350 |
+
"grad_norm": 0.09357789158821106,
|
| 1351 |
+
"learning_rate": 5.176630434782609e-06,
|
| 1352 |
+
"loss": 10.791,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.6811846689895469,
|
| 1357 |
+
"grad_norm": 0.0988670364022255,
|
| 1358 |
+
"learning_rate": 5.142663043478261e-06,
|
| 1359 |
+
"loss": 10.7917,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.6898954703832754,
|
| 1364 |
+
"grad_norm": 0.09996125102043152,
|
| 1365 |
+
"learning_rate": 5.108695652173914e-06,
|
| 1366 |
+
"loss": 10.7912,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.6986062717770034,
|
| 1371 |
+
"grad_norm": 0.09966638684272766,
|
| 1372 |
+
"learning_rate": 5.074728260869566e-06,
|
| 1373 |
+
"loss": 10.825,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.7073170731707317,
|
| 1378 |
+
"grad_norm": 0.09682171791791916,
|
| 1379 |
+
"learning_rate": 5.040760869565218e-06,
|
| 1380 |
+
"loss": 10.7906,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.71602787456446,
|
| 1385 |
+
"grad_norm": 0.09694927930831909,
|
| 1386 |
+
"learning_rate": 5.0067934782608696e-06,
|
| 1387 |
+
"loss": 10.7911,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.7247386759581882,
|
| 1392 |
+
"grad_norm": 0.09398549050092697,
|
| 1393 |
+
"learning_rate": 4.972826086956522e-06,
|
| 1394 |
+
"loss": 10.7909,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.7334494773519165,
|
| 1399 |
+
"grad_norm": 0.08475075662136078,
|
| 1400 |
+
"learning_rate": 4.938858695652175e-06,
|
| 1401 |
+
"loss": 10.7916,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.7421602787456445,
|
| 1406 |
+
"grad_norm": 0.09494160860776901,
|
| 1407 |
+
"learning_rate": 4.9048913043478265e-06,
|
| 1408 |
+
"loss": 10.7911,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.750871080139373,
|
| 1413 |
+
"grad_norm": 0.08939459174871445,
|
| 1414 |
+
"learning_rate": 4.870923913043479e-06,
|
| 1415 |
+
"loss": 10.7906,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.759581881533101,
|
| 1420 |
+
"grad_norm": 0.09646463394165039,
|
| 1421 |
+
"learning_rate": 4.836956521739131e-06,
|
| 1422 |
+
"loss": 10.7911,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.7682926829268293,
|
| 1427 |
+
"grad_norm": 0.09423504769802094,
|
| 1428 |
+
"learning_rate": 4.8029891304347835e-06,
|
| 1429 |
+
"loss": 10.7906,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.7770034843205575,
|
| 1434 |
+
"grad_norm": 0.09972324222326279,
|
| 1435 |
+
"learning_rate": 4.769021739130435e-06,
|
| 1436 |
+
"loss": 10.7914,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.7857142857142856,
|
| 1441 |
+
"grad_norm": 0.10339873284101486,
|
| 1442 |
+
"learning_rate": 4.735054347826088e-06,
|
| 1443 |
+
"loss": 10.7898,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.794425087108014,
|
| 1448 |
+
"grad_norm": 0.09187936782836914,
|
| 1449 |
+
"learning_rate": 4.7010869565217396e-06,
|
| 1450 |
+
"loss": 10.7915,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.8031358885017421,
|
| 1455 |
+
"grad_norm": 0.09024845063686371,
|
| 1456 |
+
"learning_rate": 4.667119565217391e-06,
|
| 1457 |
+
"loss": 10.7902,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.8118466898954704,
|
| 1462 |
+
"grad_norm": 0.08929532021284103,
|
| 1463 |
+
"learning_rate": 4.633152173913044e-06,
|
| 1464 |
+
"loss": 10.7905,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.8205574912891986,
|
| 1469 |
+
"grad_norm": 0.09708672016859055,
|
| 1470 |
+
"learning_rate": 4.599184782608696e-06,
|
| 1471 |
+
"loss": 10.7901,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.8292682926829267,
|
| 1476 |
+
"grad_norm": 0.10239408910274506,
|
| 1477 |
+
"learning_rate": 4.565217391304348e-06,
|
| 1478 |
+
"loss": 10.7896,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.8379790940766552,
|
| 1483 |
+
"grad_norm": 0.10148520767688751,
|
| 1484 |
+
"learning_rate": 4.53125e-06,
|
| 1485 |
+
"loss": 10.7906,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.8466898954703832,
|
| 1490 |
+
"grad_norm": 0.09579440951347351,
|
| 1491 |
+
"learning_rate": 4.497282608695653e-06,
|
| 1492 |
+
"loss": 10.7903,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.8554006968641115,
|
| 1497 |
+
"grad_norm": 0.09891035407781601,
|
| 1498 |
+
"learning_rate": 4.463315217391304e-06,
|
| 1499 |
+
"loss": 10.7904,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.8641114982578397,
|
| 1504 |
+
"grad_norm": 0.09411191940307617,
|
| 1505 |
+
"learning_rate": 4.429347826086957e-06,
|
| 1506 |
+
"loss": 10.7899,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.872822299651568,
|
| 1511 |
+
"grad_norm": 0.10932313650846481,
|
| 1512 |
+
"learning_rate": 4.395380434782609e-06,
|
| 1513 |
+
"loss": 10.7907,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.8815331010452963,
|
| 1518 |
+
"grad_norm": 0.08247391879558563,
|
| 1519 |
+
"learning_rate": 4.361413043478261e-06,
|
| 1520 |
+
"loss": 10.7902,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.8902439024390243,
|
| 1525 |
+
"grad_norm": 0.0989106297492981,
|
| 1526 |
+
"learning_rate": 4.327445652173913e-06,
|
| 1527 |
+
"loss": 10.79,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.8989547038327528,
|
| 1532 |
+
"grad_norm": 0.09022705256938934,
|
| 1533 |
+
"learning_rate": 4.293478260869566e-06,
|
| 1534 |
+
"loss": 10.7893,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.9076655052264808,
|
| 1539 |
+
"grad_norm": 0.09723115712404251,
|
| 1540 |
+
"learning_rate": 4.2595108695652174e-06,
|
| 1541 |
+
"loss": 10.7892,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.916376306620209,
|
| 1546 |
+
"grad_norm": 0.09660723060369492,
|
| 1547 |
+
"learning_rate": 4.22554347826087e-06,
|
| 1548 |
+
"loss": 10.7899,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.9250871080139373,
|
| 1553 |
+
"grad_norm": 0.09882649034261703,
|
| 1554 |
+
"learning_rate": 4.191576086956522e-06,
|
| 1555 |
+
"loss": 10.7896,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.9337979094076654,
|
| 1560 |
+
"grad_norm": 0.08623263239860535,
|
| 1561 |
+
"learning_rate": 4.157608695652174e-06,
|
| 1562 |
+
"loss": 10.7894,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.9425087108013939,
|
| 1567 |
+
"grad_norm": 0.09180337190628052,
|
| 1568 |
+
"learning_rate": 4.123641304347826e-06,
|
| 1569 |
+
"loss": 10.7896,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.951219512195122,
|
| 1574 |
+
"grad_norm": 0.09379716962575912,
|
| 1575 |
+
"learning_rate": 4.089673913043479e-06,
|
| 1576 |
+
"loss": 10.79,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.9599303135888502,
|
| 1581 |
+
"grad_norm": 0.09439699351787567,
|
| 1582 |
+
"learning_rate": 4.0557065217391305e-06,
|
| 1583 |
+
"loss": 10.79,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.9686411149825784,
|
| 1588 |
+
"grad_norm": 0.09595078229904175,
|
| 1589 |
+
"learning_rate": 4.021739130434783e-06,
|
| 1590 |
+
"loss": 10.7891,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.9773519163763065,
|
| 1595 |
+
"grad_norm": 0.09709908813238144,
|
| 1596 |
+
"learning_rate": 3.987771739130435e-06,
|
| 1597 |
+
"loss": 10.7899,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.986062717770035,
|
| 1602 |
+
"grad_norm": 0.0975862666964531,
|
| 1603 |
+
"learning_rate": 3.9538043478260875e-06,
|
| 1604 |
+
"loss": 10.7885,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.994773519163763,
|
| 1609 |
+
"grad_norm": 0.09871925413608551,
|
| 1610 |
+
"learning_rate": 3.919836956521739e-06,
|
| 1611 |
+
"loss": 10.7895,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
}
|
| 1614 |
+
],
|
| 1615 |
+
"logging_steps": 10,
|
| 1616 |
+
"max_steps": 3444,
|
| 1617 |
+
"num_input_tokens_seen": 0,
|
| 1618 |
+
"num_train_epochs": 3,
|
| 1619 |
+
"save_steps": 500,
|
| 1620 |
+
"stateful_callbacks": {
|
| 1621 |
+
"TrainerControl": {
|
| 1622 |
+
"args": {
|
| 1623 |
+
"should_epoch_stop": false,
|
| 1624 |
+
"should_evaluate": false,
|
| 1625 |
+
"should_log": false,
|
| 1626 |
+
"should_save": true,
|
| 1627 |
+
"should_training_stop": false
|
| 1628 |
+
},
|
| 1629 |
+
"attributes": {}
|
| 1630 |
+
}
|
| 1631 |
+
},
|
| 1632 |
+
"total_flos": 1131725678628864.0,
|
| 1633 |
+
"train_batch_size": 32,
|
| 1634 |
+
"trial_name": null,
|
| 1635 |
+
"trial_params": null
|
| 1636 |
+
}
|
checkpoint-2296/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90c67833faf31d4402ee0893264f8627505d9d95e6510f92d119f2d25dc10219
|
| 3 |
+
size 5176
|
checkpoint-2296/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3444/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "C:/Users/wonde/Wonder-Griffin/TraXLMistralForCausalLM",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TraXLMistralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"dropout": 0.1,
|
| 7 |
+
"dynamic_routing": true,
|
| 8 |
+
"ff_expansion_factor": 4,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "LABEL_0",
|
| 12 |
+
"1": "LABEL_1",
|
| 13 |
+
"2": "LABEL_2",
|
| 14 |
+
"3": "LABEL_3",
|
| 15 |
+
"4": "LABEL_4"
|
| 16 |
+
},
|
| 17 |
+
"is_decoder": true,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"LABEL_0": 0,
|
| 20 |
+
"LABEL_1": 1,
|
| 21 |
+
"LABEL_2": 2,
|
| 22 |
+
"LABEL_3": 3,
|
| 23 |
+
"LABEL_4": 4
|
| 24 |
+
},
|
| 25 |
+
"max_computation_steps": 5,
|
| 26 |
+
"max_len": 256,
|
| 27 |
+
"memory_size": 256,
|
| 28 |
+
"model_type": "TraXLMistral",
|
| 29 |
+
"n_head": 4,
|
| 30 |
+
"n_layer": 4,
|
| 31 |
+
"rnn_units": 128,
|
| 32 |
+
"sparse_attention": true,
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.44.2",
|
| 35 |
+
"vocab_size": 50257
|
| 36 |
+
}
|
checkpoint-3444/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3444/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0bceb990b83c77292376f2a509df9fffdf04ccd9c395f19d8d148acfaa547d1
|
| 3 |
+
size 194528200
|
checkpoint-3444/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:393320e81d046e5f0d09750967fb701b4b53934fd6c289d09d2803abbc0f3468
|
| 3 |
+
size 369911930
|
checkpoint-3444/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35f49f4e5b8e98b628d31fbfd7483c08a39a1b948cada64044dd1dc501eb00ae
|
| 3 |
+
size 14244
|
checkpoint-3444/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:671d5ca27acbe7011faba7a4eee568d11ed581975ccc473a52d955b3ece09bd4
|
| 3 |
+
size 1064
|
checkpoint-3444/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
checkpoint-3444/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3444/tokenizer_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"bos_token": "<|endoftext|>",
|
| 14 |
+
"clean_up_tokenization_spaces": true,
|
| 15 |
+
"eos_token": "<|endoftext|>",
|
| 16 |
+
"model_max_length": 1024,
|
| 17 |
+
"pad_token": "<|endoftext|>",
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>"
|
| 20 |
+
}
|
checkpoint-3444/trainer_state.json
ADDED
|
@@ -0,0 +1,2441 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3444,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008710801393728223,
|
| 13 |
+
"grad_norm": 96.23190307617188,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 10.808,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.017421602787456445,
|
| 20 |
+
"grad_norm": 463.5845642089844,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 10.8147,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02613240418118467,
|
| 27 |
+
"grad_norm": 0.10578063130378723,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 10.8077,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03484320557491289,
|
| 34 |
+
"grad_norm": 0.09780663996934891,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.8218,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04355400696864112,
|
| 41 |
+
"grad_norm": 1469.5677490234375,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.8079,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05226480836236934,
|
| 48 |
+
"grad_norm": 0.09817048907279968,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 10.8204,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06097560975609756,
|
| 55 |
+
"grad_norm": 0.09422990679740906,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 10.8159,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06968641114982578,
|
| 62 |
+
"grad_norm": 0.09545309096574783,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 10.811,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.078397212543554,
|
| 69 |
+
"grad_norm": 0.08887256681919098,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 10.8048,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08710801393728224,
|
| 76 |
+
"grad_norm": 0.0909484326839447,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 10.8126,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09581881533101046,
|
| 83 |
+
"grad_norm": 0.09697296470403671,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 10.8056,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10452961672473868,
|
| 90 |
+
"grad_norm": 0.09927204251289368,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 11.0541,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1132404181184669,
|
| 97 |
+
"grad_norm": 0.09392143785953522,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 10.8073,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.12195121951219512,
|
| 104 |
+
"grad_norm": 0.09451345354318619,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 10.8041,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.13066202090592335,
|
| 111 |
+
"grad_norm": 0.09947695583105087,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 10.8141,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13937282229965156,
|
| 118 |
+
"grad_norm": 0.09803537279367447,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 10.805,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1480836236933798,
|
| 125 |
+
"grad_norm": 0.09703430533409119,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 10.8193,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.156794425087108,
|
| 132 |
+
"grad_norm": 0.09272942692041397,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 10.8151,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16550522648083624,
|
| 139 |
+
"grad_norm": 0.0950935110449791,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 10.8051,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.17421602787456447,
|
| 146 |
+
"grad_norm": 0.10270467400550842,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 10.8043,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.18292682926829268,
|
| 153 |
+
"grad_norm": 0.09881923347711563,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 10.8044,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1916376306620209,
|
| 160 |
+
"grad_norm": 0.10348394513130188,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 10.8036,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.20034843205574912,
|
| 167 |
+
"grad_norm": 0.09483418613672256,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 10.8046,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20905923344947736,
|
| 174 |
+
"grad_norm": 0.09620082378387451,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 10.807,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21777003484320556,
|
| 181 |
+
"grad_norm": 0.09709980338811874,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 10.8058,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2264808362369338,
|
| 188 |
+
"grad_norm": 0.09373420476913452,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 10.8044,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.23519163763066203,
|
| 195 |
+
"grad_norm": 0.09700135141611099,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 10.8168,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.24390243902439024,
|
| 202 |
+
"grad_norm": 0.09450981020927429,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 10.8043,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.25261324041811845,
|
| 209 |
+
"grad_norm": 0.09983480721712112,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 10.8037,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2613240418118467,
|
| 216 |
+
"grad_norm": 0.09153986722230911,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 10.8041,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2700348432055749,
|
| 223 |
+
"grad_norm": 0.09664902836084366,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 10.805,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2787456445993031,
|
| 230 |
+
"grad_norm": 0.08636524528265,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 10.8049,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2874564459930314,
|
| 237 |
+
"grad_norm": 0.0945158526301384,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 10.8047,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2961672473867596,
|
| 244 |
+
"grad_norm": 0.09447965025901794,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 10.8044,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3048780487804878,
|
| 251 |
+
"grad_norm": 0.0949367955327034,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 10.8038,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.313588850174216,
|
| 258 |
+
"grad_norm": 0.10188119113445282,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 10.8037,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.32229965156794427,
|
| 265 |
+
"grad_norm": 0.09333271533250809,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 10.8029,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3310104529616725,
|
| 272 |
+
"grad_norm": 0.10046978294849396,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 10.8037,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3397212543554007,
|
| 279 |
+
"grad_norm": 0.09508766233921051,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 10.8065,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.34843205574912894,
|
| 286 |
+
"grad_norm": 0.09448560327291489,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 10.8058,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.35714285714285715,
|
| 293 |
+
"grad_norm": 0.09986421465873718,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 14.4299,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.36585365853658536,
|
| 300 |
+
"grad_norm": 0.10298220068216324,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 10.8035,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.37456445993031356,
|
| 307 |
+
"grad_norm": 0.09236953407526016,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 10.8309,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3832752613240418,
|
| 314 |
+
"grad_norm": 0.09000707417726517,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 10.8027,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.39198606271777003,
|
| 321 |
+
"grad_norm": 0.09554922580718994,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 10.8045,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.40069686411149824,
|
| 328 |
+
"grad_norm": 130.35623168945312,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 10.8034,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.4094076655052265,
|
| 335 |
+
"grad_norm": 0.09580319374799728,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 10.8022,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4181184668989547,
|
| 342 |
+
"grad_norm": 0.10328342765569687,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 10.8087,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4268292682926829,
|
| 349 |
+
"grad_norm": 0.09383571892976761,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 10.8027,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4355400696864111,
|
| 356 |
+
"grad_norm": 61532.49609375,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 15.4887,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4442508710801394,
|
| 363 |
+
"grad_norm": 0.09285067766904831,
|
| 364 |
+
"learning_rate": 9.966032608695653e-06,
|
| 365 |
+
"loss": 10.8028,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4529616724738676,
|
| 370 |
+
"grad_norm": 0.09329842031002045,
|
| 371 |
+
"learning_rate": 9.932065217391306e-06,
|
| 372 |
+
"loss": 10.8969,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4616724738675958,
|
| 377 |
+
"grad_norm": 0.09399975091218948,
|
| 378 |
+
"learning_rate": 9.898097826086957e-06,
|
| 379 |
+
"loss": 10.8297,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.47038327526132406,
|
| 384 |
+
"grad_norm": 1130.7960205078125,
|
| 385 |
+
"learning_rate": 9.86413043478261e-06,
|
| 386 |
+
"loss": 10.8358,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.47909407665505227,
|
| 391 |
+
"grad_norm": 0.0971442461013794,
|
| 392 |
+
"learning_rate": 9.830163043478262e-06,
|
| 393 |
+
"loss": 10.8641,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4878048780487805,
|
| 398 |
+
"grad_norm": 0.09969327598810196,
|
| 399 |
+
"learning_rate": 9.796195652173915e-06,
|
| 400 |
+
"loss": 10.8124,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4965156794425087,
|
| 405 |
+
"grad_norm": 0.0888582393527031,
|
| 406 |
+
"learning_rate": 9.762228260869566e-06,
|
| 407 |
+
"loss": 10.9576,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5052264808362369,
|
| 412 |
+
"grad_norm": 0.09847419708967209,
|
| 413 |
+
"learning_rate": 9.728260869565218e-06,
|
| 414 |
+
"loss": 10.8022,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5139372822299652,
|
| 419 |
+
"grad_norm": 0.0947929099202156,
|
| 420 |
+
"learning_rate": 9.694293478260869e-06,
|
| 421 |
+
"loss": 10.8011,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5226480836236934,
|
| 426 |
+
"grad_norm": 9.681638717651367,
|
| 427 |
+
"learning_rate": 9.660326086956523e-06,
|
| 428 |
+
"loss": 10.8015,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5313588850174216,
|
| 433 |
+
"grad_norm": 0.09560558199882507,
|
| 434 |
+
"learning_rate": 9.626358695652174e-06,
|
| 435 |
+
"loss": 13.2261,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5400696864111498,
|
| 440 |
+
"grad_norm": 0.09392021596431732,
|
| 441 |
+
"learning_rate": 9.592391304347827e-06,
|
| 442 |
+
"loss": 10.802,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5487804878048781,
|
| 447 |
+
"grad_norm": 0.10044202208518982,
|
| 448 |
+
"learning_rate": 9.558423913043478e-06,
|
| 449 |
+
"loss": 10.8005,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5574912891986062,
|
| 454 |
+
"grad_norm": 0.09005896002054214,
|
| 455 |
+
"learning_rate": 9.524456521739132e-06,
|
| 456 |
+
"loss": 10.802,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5662020905923345,
|
| 461 |
+
"grad_norm": 0.09686450660228729,
|
| 462 |
+
"learning_rate": 9.490489130434783e-06,
|
| 463 |
+
"loss": 10.801,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5749128919860628,
|
| 468 |
+
"grad_norm": 0.09321983903646469,
|
| 469 |
+
"learning_rate": 9.456521739130436e-06,
|
| 470 |
+
"loss": 10.8008,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5836236933797909,
|
| 475 |
+
"grad_norm": 0.0884283185005188,
|
| 476 |
+
"learning_rate": 9.422554347826087e-06,
|
| 477 |
+
"loss": 10.801,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5923344947735192,
|
| 482 |
+
"grad_norm": 0.09928812086582184,
|
| 483 |
+
"learning_rate": 9.388586956521741e-06,
|
| 484 |
+
"loss": 10.7998,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6010452961672473,
|
| 489 |
+
"grad_norm": 0.09434370696544647,
|
| 490 |
+
"learning_rate": 9.354619565217392e-06,
|
| 491 |
+
"loss": 10.8007,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6097560975609756,
|
| 496 |
+
"grad_norm": 0.09228064864873886,
|
| 497 |
+
"learning_rate": 9.320652173913044e-06,
|
| 498 |
+
"loss": 10.8002,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6184668989547039,
|
| 503 |
+
"grad_norm": 0.09412040561437607,
|
| 504 |
+
"learning_rate": 9.286684782608695e-06,
|
| 505 |
+
"loss": 10.8009,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.627177700348432,
|
| 510 |
+
"grad_norm": 0.0975504145026207,
|
| 511 |
+
"learning_rate": 9.25271739130435e-06,
|
| 512 |
+
"loss": 10.8002,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6358885017421603,
|
| 517 |
+
"grad_norm": 0.10332240164279938,
|
| 518 |
+
"learning_rate": 9.21875e-06,
|
| 519 |
+
"loss": 10.8001,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6445993031358885,
|
| 524 |
+
"grad_norm": 0.09318797290325165,
|
| 525 |
+
"learning_rate": 9.184782608695653e-06,
|
| 526 |
+
"loss": 10.8001,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6533101045296167,
|
| 531 |
+
"grad_norm": 0.09171754121780396,
|
| 532 |
+
"learning_rate": 9.150815217391306e-06,
|
| 533 |
+
"loss": 10.7999,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.662020905923345,
|
| 538 |
+
"grad_norm": 0.0921366959810257,
|
| 539 |
+
"learning_rate": 9.116847826086958e-06,
|
| 540 |
+
"loss": 10.8003,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6707317073170732,
|
| 545 |
+
"grad_norm": 0.09661433100700378,
|
| 546 |
+
"learning_rate": 9.08288043478261e-06,
|
| 547 |
+
"loss": 10.8001,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6794425087108014,
|
| 552 |
+
"grad_norm": 0.09323884546756744,
|
| 553 |
+
"learning_rate": 9.048913043478262e-06,
|
| 554 |
+
"loss": 10.7997,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6881533101045296,
|
| 559 |
+
"grad_norm": 0.08752495795488358,
|
| 560 |
+
"learning_rate": 9.014945652173914e-06,
|
| 561 |
+
"loss": 10.8,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6968641114982579,
|
| 566 |
+
"grad_norm": 0.09087631851434708,
|
| 567 |
+
"learning_rate": 8.980978260869567e-06,
|
| 568 |
+
"loss": 10.798,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.705574912891986,
|
| 573 |
+
"grad_norm": 0.09523475170135498,
|
| 574 |
+
"learning_rate": 8.947010869565218e-06,
|
| 575 |
+
"loss": 10.7994,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7142857142857143,
|
| 580 |
+
"grad_norm": 0.09918010234832764,
|
| 581 |
+
"learning_rate": 8.91304347826087e-06,
|
| 582 |
+
"loss": 10.7996,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.7229965156794426,
|
| 587 |
+
"grad_norm": 0.09539210051298141,
|
| 588 |
+
"learning_rate": 8.879076086956523e-06,
|
| 589 |
+
"loss": 10.7993,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.7317073170731707,
|
| 594 |
+
"grad_norm": 0.09296409040689468,
|
| 595 |
+
"learning_rate": 8.845108695652174e-06,
|
| 596 |
+
"loss": 10.7993,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.740418118466899,
|
| 601 |
+
"grad_norm": 0.08845611661672592,
|
| 602 |
+
"learning_rate": 8.811141304347827e-06,
|
| 603 |
+
"loss": 10.7993,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.7491289198606271,
|
| 608 |
+
"grad_norm": 0.0954870954155922,
|
| 609 |
+
"learning_rate": 8.77717391304348e-06,
|
| 610 |
+
"loss": 10.7981,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7578397212543554,
|
| 615 |
+
"grad_norm": 0.09371493011713028,
|
| 616 |
+
"learning_rate": 8.743206521739132e-06,
|
| 617 |
+
"loss": 10.7984,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7665505226480837,
|
| 622 |
+
"grad_norm": 0.09903618693351746,
|
| 623 |
+
"learning_rate": 8.709239130434783e-06,
|
| 624 |
+
"loss": 10.7987,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7752613240418118,
|
| 629 |
+
"grad_norm": 0.08991510421037674,
|
| 630 |
+
"learning_rate": 8.675271739130435e-06,
|
| 631 |
+
"loss": 10.7976,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7839721254355401,
|
| 636 |
+
"grad_norm": 0.09141765534877777,
|
| 637 |
+
"learning_rate": 8.641304347826088e-06,
|
| 638 |
+
"loss": 10.799,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7926829268292683,
|
| 643 |
+
"grad_norm": 0.09670023620128632,
|
| 644 |
+
"learning_rate": 8.60733695652174e-06,
|
| 645 |
+
"loss": 10.7973,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8013937282229965,
|
| 650 |
+
"grad_norm": 0.09337509423494339,
|
| 651 |
+
"learning_rate": 8.573369565217391e-06,
|
| 652 |
+
"loss": 10.7985,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.8101045296167247,
|
| 657 |
+
"grad_norm": 0.09618163853883743,
|
| 658 |
+
"learning_rate": 8.539402173913044e-06,
|
| 659 |
+
"loss": 10.7989,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.818815331010453,
|
| 664 |
+
"grad_norm": 0.09537822753190994,
|
| 665 |
+
"learning_rate": 8.505434782608697e-06,
|
| 666 |
+
"loss": 10.7981,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.8275261324041812,
|
| 671 |
+
"grad_norm": 0.09312313050031662,
|
| 672 |
+
"learning_rate": 8.47146739130435e-06,
|
| 673 |
+
"loss": 10.7984,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.8362369337979094,
|
| 678 |
+
"grad_norm": 0.09118187427520752,
|
| 679 |
+
"learning_rate": 8.4375e-06,
|
| 680 |
+
"loss": 10.7981,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.8449477351916377,
|
| 685 |
+
"grad_norm": 0.0919727087020874,
|
| 686 |
+
"learning_rate": 8.403532608695653e-06,
|
| 687 |
+
"loss": 10.7978,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8536585365853658,
|
| 692 |
+
"grad_norm": 0.09706488996744156,
|
| 693 |
+
"learning_rate": 8.369565217391305e-06,
|
| 694 |
+
"loss": 10.7979,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8623693379790941,
|
| 699 |
+
"grad_norm": 0.0944119319319725,
|
| 700 |
+
"learning_rate": 8.335597826086958e-06,
|
| 701 |
+
"loss": 10.7983,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8710801393728222,
|
| 706 |
+
"grad_norm": 0.1016387864947319,
|
| 707 |
+
"learning_rate": 8.301630434782609e-06,
|
| 708 |
+
"loss": 10.7973,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8797909407665505,
|
| 713 |
+
"grad_norm": 0.10198845714330673,
|
| 714 |
+
"learning_rate": 8.267663043478261e-06,
|
| 715 |
+
"loss": 10.7984,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8885017421602788,
|
| 720 |
+
"grad_norm": 0.10102390497922897,
|
| 721 |
+
"learning_rate": 8.233695652173914e-06,
|
| 722 |
+
"loss": 10.7971,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8972125435540069,
|
| 727 |
+
"grad_norm": 0.0972791463136673,
|
| 728 |
+
"learning_rate": 8.199728260869567e-06,
|
| 729 |
+
"loss": 10.7971,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9059233449477352,
|
| 734 |
+
"grad_norm": 0.09343789517879486,
|
| 735 |
+
"learning_rate": 8.165760869565218e-06,
|
| 736 |
+
"loss": 10.7978,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.9146341463414634,
|
| 741 |
+
"grad_norm": 0.09384151548147202,
|
| 742 |
+
"learning_rate": 8.13179347826087e-06,
|
| 743 |
+
"loss": 10.797,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.9233449477351916,
|
| 748 |
+
"grad_norm": 0.09324201196432114,
|
| 749 |
+
"learning_rate": 8.097826086956523e-06,
|
| 750 |
+
"loss": 10.7969,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.9320557491289199,
|
| 755 |
+
"grad_norm": 0.09341388940811157,
|
| 756 |
+
"learning_rate": 8.063858695652175e-06,
|
| 757 |
+
"loss": 10.7972,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9407665505226481,
|
| 762 |
+
"grad_norm": 0.09563016146421432,
|
| 763 |
+
"learning_rate": 8.029891304347826e-06,
|
| 764 |
+
"loss": 10.7978,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9494773519163763,
|
| 769 |
+
"grad_norm": 0.09752058237791061,
|
| 770 |
+
"learning_rate": 7.995923913043479e-06,
|
| 771 |
+
"loss": 10.7962,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9581881533101045,
|
| 776 |
+
"grad_norm": 0.09652635455131531,
|
| 777 |
+
"learning_rate": 7.961956521739131e-06,
|
| 778 |
+
"loss": 10.7968,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9668989547038328,
|
| 783 |
+
"grad_norm": 0.097730852663517,
|
| 784 |
+
"learning_rate": 7.927989130434784e-06,
|
| 785 |
+
"loss": 10.7967,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.975609756097561,
|
| 790 |
+
"grad_norm": 0.09178122133016586,
|
| 791 |
+
"learning_rate": 7.894021739130435e-06,
|
| 792 |
+
"loss": 10.7963,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9843205574912892,
|
| 797 |
+
"grad_norm": 0.09718208760023117,
|
| 798 |
+
"learning_rate": 7.860054347826088e-06,
|
| 799 |
+
"loss": 10.7967,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9930313588850174,
|
| 804 |
+
"grad_norm": 0.09228435158729553,
|
| 805 |
+
"learning_rate": 7.82608695652174e-06,
|
| 806 |
+
"loss": 10.7973,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.0017421602787457,
|
| 811 |
+
"grad_norm": 0.08885621279478073,
|
| 812 |
+
"learning_rate": 7.792119565217393e-06,
|
| 813 |
+
"loss": 10.796,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.0104529616724738,
|
| 818 |
+
"grad_norm": 0.09165474027395248,
|
| 819 |
+
"learning_rate": 7.758152173913044e-06,
|
| 820 |
+
"loss": 10.7957,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.019163763066202,
|
| 825 |
+
"grad_norm": 0.09650120139122009,
|
| 826 |
+
"learning_rate": 7.724184782608696e-06,
|
| 827 |
+
"loss": 10.7967,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.0278745644599303,
|
| 832 |
+
"grad_norm": 0.10563351958990097,
|
| 833 |
+
"learning_rate": 7.690217391304349e-06,
|
| 834 |
+
"loss": 10.7963,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.0365853658536586,
|
| 839 |
+
"grad_norm": 0.09407645463943481,
|
| 840 |
+
"learning_rate": 7.656250000000001e-06,
|
| 841 |
+
"loss": 10.7958,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.0452961672473868,
|
| 846 |
+
"grad_norm": 0.100870780646801,
|
| 847 |
+
"learning_rate": 7.622282608695652e-06,
|
| 848 |
+
"loss": 10.7961,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.054006968641115,
|
| 853 |
+
"grad_norm": 0.09036817401647568,
|
| 854 |
+
"learning_rate": 7.588315217391306e-06,
|
| 855 |
+
"loss": 10.7965,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.0627177700348431,
|
| 860 |
+
"grad_norm": 0.09792335331439972,
|
| 861 |
+
"learning_rate": 7.5543478260869576e-06,
|
| 862 |
+
"loss": 10.7951,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.0714285714285714,
|
| 867 |
+
"grad_norm": 0.10280462354421616,
|
| 868 |
+
"learning_rate": 7.520380434782609e-06,
|
| 869 |
+
"loss": 10.796,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0801393728222997,
|
| 874 |
+
"grad_norm": 0.09284704178571701,
|
| 875 |
+
"learning_rate": 7.486413043478261e-06,
|
| 876 |
+
"loss": 10.796,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.088850174216028,
|
| 881 |
+
"grad_norm": 0.09495861083269119,
|
| 882 |
+
"learning_rate": 7.4524456521739145e-06,
|
| 883 |
+
"loss": 10.7954,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.0975609756097562,
|
| 888 |
+
"grad_norm": 0.0963548794388771,
|
| 889 |
+
"learning_rate": 7.418478260869566e-06,
|
| 890 |
+
"loss": 10.7954,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.1062717770034842,
|
| 895 |
+
"grad_norm": 0.10302776098251343,
|
| 896 |
+
"learning_rate": 7.384510869565218e-06,
|
| 897 |
+
"loss": 10.7946,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.1149825783972125,
|
| 902 |
+
"grad_norm": 0.09135697036981583,
|
| 903 |
+
"learning_rate": 7.35054347826087e-06,
|
| 904 |
+
"loss": 10.7951,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.1236933797909407,
|
| 909 |
+
"grad_norm": 0.09077905118465424,
|
| 910 |
+
"learning_rate": 7.316576086956522e-06,
|
| 911 |
+
"loss": 10.7949,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.132404181184669,
|
| 916 |
+
"grad_norm": 0.09623997658491135,
|
| 917 |
+
"learning_rate": 7.282608695652175e-06,
|
| 918 |
+
"loss": 10.795,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.1411149825783973,
|
| 923 |
+
"grad_norm": 0.09430749714374542,
|
| 924 |
+
"learning_rate": 7.248641304347827e-06,
|
| 925 |
+
"loss": 10.7954,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.1498257839721253,
|
| 930 |
+
"grad_norm": 0.09416501224040985,
|
| 931 |
+
"learning_rate": 7.2146739130434785e-06,
|
| 932 |
+
"loss": 10.795,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.1585365853658536,
|
| 937 |
+
"grad_norm": 0.09839834272861481,
|
| 938 |
+
"learning_rate": 7.180706521739131e-06,
|
| 939 |
+
"loss": 10.795,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.1672473867595818,
|
| 944 |
+
"grad_norm": 0.09414810687303543,
|
| 945 |
+
"learning_rate": 7.146739130434784e-06,
|
| 946 |
+
"loss": 10.7949,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.17595818815331,
|
| 951 |
+
"grad_norm": 0.09453831613063812,
|
| 952 |
+
"learning_rate": 7.1127717391304354e-06,
|
| 953 |
+
"loss": 10.942,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.1846689895470384,
|
| 958 |
+
"grad_norm": 637.3465576171875,
|
| 959 |
+
"learning_rate": 7.078804347826087e-06,
|
| 960 |
+
"loss": 10.8174,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.1933797909407666,
|
| 965 |
+
"grad_norm": 0.09487204998731613,
|
| 966 |
+
"learning_rate": 7.04483695652174e-06,
|
| 967 |
+
"loss": 16.214,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.202090592334495,
|
| 972 |
+
"grad_norm": 0.08683289587497711,
|
| 973 |
+
"learning_rate": 7.0108695652173915e-06,
|
| 974 |
+
"loss": 10.7961,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.210801393728223,
|
| 979 |
+
"grad_norm": 0.09138838201761246,
|
| 980 |
+
"learning_rate": 6.976902173913044e-06,
|
| 981 |
+
"loss": 10.7954,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.2195121951219512,
|
| 986 |
+
"grad_norm": 0.09047773480415344,
|
| 987 |
+
"learning_rate": 6.942934782608696e-06,
|
| 988 |
+
"loss": 10.7941,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.2282229965156795,
|
| 993 |
+
"grad_norm": 0.10373370349407196,
|
| 994 |
+
"learning_rate": 6.9089673913043485e-06,
|
| 995 |
+
"loss": 10.7933,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.2369337979094077,
|
| 1000 |
+
"grad_norm": 0.0910143107175827,
|
| 1001 |
+
"learning_rate": 6.875e-06,
|
| 1002 |
+
"loss": 10.7952,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.245644599303136,
|
| 1007 |
+
"grad_norm": 0.08921236544847488,
|
| 1008 |
+
"learning_rate": 6.841032608695653e-06,
|
| 1009 |
+
"loss": 10.7945,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.254355400696864,
|
| 1014 |
+
"grad_norm": 0.09373684972524643,
|
| 1015 |
+
"learning_rate": 6.8070652173913054e-06,
|
| 1016 |
+
"loss": 10.795,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.2630662020905923,
|
| 1021 |
+
"grad_norm": 0.09562421590089798,
|
| 1022 |
+
"learning_rate": 6.773097826086957e-06,
|
| 1023 |
+
"loss": 10.7948,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.2717770034843205,
|
| 1028 |
+
"grad_norm": 0.0909007340669632,
|
| 1029 |
+
"learning_rate": 6.739130434782609e-06,
|
| 1030 |
+
"loss": 10.7945,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.2804878048780488,
|
| 1035 |
+
"grad_norm": 0.09700188040733337,
|
| 1036 |
+
"learning_rate": 6.705163043478261e-06,
|
| 1037 |
+
"loss": 10.7947,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.289198606271777,
|
| 1042 |
+
"grad_norm": 0.09472058713436127,
|
| 1043 |
+
"learning_rate": 6.671195652173914e-06,
|
| 1044 |
+
"loss": 10.7942,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.297909407665505,
|
| 1049 |
+
"grad_norm": 0.09371300041675568,
|
| 1050 |
+
"learning_rate": 6.637228260869566e-06,
|
| 1051 |
+
"loss": 10.793,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.3066202090592334,
|
| 1056 |
+
"grad_norm": 0.08980853110551834,
|
| 1057 |
+
"learning_rate": 6.603260869565218e-06,
|
| 1058 |
+
"loss": 10.7936,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.3153310104529616,
|
| 1063 |
+
"grad_norm": 0.09769923985004425,
|
| 1064 |
+
"learning_rate": 6.569293478260869e-06,
|
| 1065 |
+
"loss": 10.7939,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.32404181184669,
|
| 1070 |
+
"grad_norm": 0.0956735908985138,
|
| 1071 |
+
"learning_rate": 6.535326086956523e-06,
|
| 1072 |
+
"loss": 10.7934,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.3327526132404182,
|
| 1077 |
+
"grad_norm": 0.09061311185359955,
|
| 1078 |
+
"learning_rate": 6.501358695652175e-06,
|
| 1079 |
+
"loss": 10.7937,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.3414634146341464,
|
| 1084 |
+
"grad_norm": 72412.171875,
|
| 1085 |
+
"learning_rate": 6.467391304347826e-06,
|
| 1086 |
+
"loss": 13.1839,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.3501742160278747,
|
| 1091 |
+
"grad_norm": 0.09127330780029297,
|
| 1092 |
+
"learning_rate": 6.433423913043478e-06,
|
| 1093 |
+
"loss": 13.2599,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.3588850174216027,
|
| 1098 |
+
"grad_norm": 0.09337225556373596,
|
| 1099 |
+
"learning_rate": 6.3994565217391316e-06,
|
| 1100 |
+
"loss": 10.8129,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.367595818815331,
|
| 1105 |
+
"grad_norm": 0.10567159950733185,
|
| 1106 |
+
"learning_rate": 6.365489130434783e-06,
|
| 1107 |
+
"loss": 10.7932,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3763066202090593,
|
| 1112 |
+
"grad_norm": 0.09582757949829102,
|
| 1113 |
+
"learning_rate": 6.331521739130435e-06,
|
| 1114 |
+
"loss": 10.7934,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.3850174216027875,
|
| 1119 |
+
"grad_norm": 0.09011757373809814,
|
| 1120 |
+
"learning_rate": 6.297554347826087e-06,
|
| 1121 |
+
"loss": 10.7934,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.3937282229965158,
|
| 1126 |
+
"grad_norm": 0.09461648017168045,
|
| 1127 |
+
"learning_rate": 6.26358695652174e-06,
|
| 1128 |
+
"loss": 10.7922,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.4024390243902438,
|
| 1133 |
+
"grad_norm": 0.0957835465669632,
|
| 1134 |
+
"learning_rate": 6.229619565217392e-06,
|
| 1135 |
+
"loss": 10.793,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.411149825783972,
|
| 1140 |
+
"grad_norm": 0.09626618772745132,
|
| 1141 |
+
"learning_rate": 6.195652173913044e-06,
|
| 1142 |
+
"loss": 10.7918,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.4198606271777003,
|
| 1147 |
+
"grad_norm": 0.09974712133407593,
|
| 1148 |
+
"learning_rate": 6.1616847826086955e-06,
|
| 1149 |
+
"loss": 10.7932,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.4285714285714286,
|
| 1154 |
+
"grad_norm": 0.09606435894966125,
|
| 1155 |
+
"learning_rate": 6.127717391304349e-06,
|
| 1156 |
+
"loss": 10.7926,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.4372822299651569,
|
| 1161 |
+
"grad_norm": 0.10175999999046326,
|
| 1162 |
+
"learning_rate": 6.093750000000001e-06,
|
| 1163 |
+
"loss": 10.7929,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.445993031358885,
|
| 1168 |
+
"grad_norm": 0.09452734142541885,
|
| 1169 |
+
"learning_rate": 6.0597826086956525e-06,
|
| 1170 |
+
"loss": 10.7936,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.4547038327526132,
|
| 1175 |
+
"grad_norm": 0.09544497728347778,
|
| 1176 |
+
"learning_rate": 6.025815217391305e-06,
|
| 1177 |
+
"loss": 10.793,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.4634146341463414,
|
| 1182 |
+
"grad_norm": 0.0867006853222847,
|
| 1183 |
+
"learning_rate": 5.991847826086957e-06,
|
| 1184 |
+
"loss": 10.7931,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.4721254355400697,
|
| 1189 |
+
"grad_norm": 0.09889491647481918,
|
| 1190 |
+
"learning_rate": 5.9578804347826094e-06,
|
| 1191 |
+
"loss": 10.7926,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.480836236933798,
|
| 1196 |
+
"grad_norm": 0.09840565174818039,
|
| 1197 |
+
"learning_rate": 5.923913043478261e-06,
|
| 1198 |
+
"loss": 10.7926,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.489547038327526,
|
| 1203 |
+
"grad_norm": 0.09129170328378677,
|
| 1204 |
+
"learning_rate": 5.889945652173914e-06,
|
| 1205 |
+
"loss": 10.7921,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.4982578397212545,
|
| 1210 |
+
"grad_norm": 0.09560216963291168,
|
| 1211 |
+
"learning_rate": 5.8559782608695656e-06,
|
| 1212 |
+
"loss": 10.7924,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.5069686411149825,
|
| 1217 |
+
"grad_norm": 0.09289251267910004,
|
| 1218 |
+
"learning_rate": 5.822010869565218e-06,
|
| 1219 |
+
"loss": 10.792,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.5156794425087108,
|
| 1224 |
+
"grad_norm": 0.09835352748632431,
|
| 1225 |
+
"learning_rate": 5.78804347826087e-06,
|
| 1226 |
+
"loss": 10.7917,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.524390243902439,
|
| 1231 |
+
"grad_norm": 0.10089673846960068,
|
| 1232 |
+
"learning_rate": 5.7540760869565225e-06,
|
| 1233 |
+
"loss": 10.7921,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.533101045296167,
|
| 1238 |
+
"grad_norm": 0.0979476273059845,
|
| 1239 |
+
"learning_rate": 5.720108695652174e-06,
|
| 1240 |
+
"loss": 10.7933,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.5418118466898956,
|
| 1245 |
+
"grad_norm": 0.10268489271402359,
|
| 1246 |
+
"learning_rate": 5.686141304347826e-06,
|
| 1247 |
+
"loss": 10.7946,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.5505226480836236,
|
| 1252 |
+
"grad_norm": 0.09039770066738129,
|
| 1253 |
+
"learning_rate": 5.652173913043479e-06,
|
| 1254 |
+
"loss": 10.7923,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.5592334494773519,
|
| 1259 |
+
"grad_norm": 0.09356367588043213,
|
| 1260 |
+
"learning_rate": 5.618206521739131e-06,
|
| 1261 |
+
"loss": 10.7918,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.5679442508710801,
|
| 1266 |
+
"grad_norm": 0.09169076383113861,
|
| 1267 |
+
"learning_rate": 5.584239130434783e-06,
|
| 1268 |
+
"loss": 10.7919,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.5766550522648084,
|
| 1273 |
+
"grad_norm": 0.09374472498893738,
|
| 1274 |
+
"learning_rate": 5.550271739130435e-06,
|
| 1275 |
+
"loss": 10.7922,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.5853658536585367,
|
| 1280 |
+
"grad_norm": 0.09261377900838852,
|
| 1281 |
+
"learning_rate": 5.516304347826087e-06,
|
| 1282 |
+
"loss": 10.7911,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.5940766550522647,
|
| 1287 |
+
"grad_norm": 0.0917358323931694,
|
| 1288 |
+
"learning_rate": 5.48233695652174e-06,
|
| 1289 |
+
"loss": 10.7926,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.6027874564459932,
|
| 1294 |
+
"grad_norm": 0.09259490668773651,
|
| 1295 |
+
"learning_rate": 5.448369565217392e-06,
|
| 1296 |
+
"loss": 10.7913,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.6114982578397212,
|
| 1301 |
+
"grad_norm": 0.09436757117509842,
|
| 1302 |
+
"learning_rate": 5.4144021739130434e-06,
|
| 1303 |
+
"loss": 10.7907,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.6202090592334495,
|
| 1308 |
+
"grad_norm": 0.0956592783331871,
|
| 1309 |
+
"learning_rate": 5.380434782608695e-06,
|
| 1310 |
+
"loss": 10.7925,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.6289198606271778,
|
| 1315 |
+
"grad_norm": 0.09467855840921402,
|
| 1316 |
+
"learning_rate": 5.346467391304349e-06,
|
| 1317 |
+
"loss": 10.7921,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.6376306620209058,
|
| 1322 |
+
"grad_norm": 0.0989622175693512,
|
| 1323 |
+
"learning_rate": 5.3125e-06,
|
| 1324 |
+
"loss": 10.7918,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.6463414634146343,
|
| 1329 |
+
"grad_norm": 0.09564294666051865,
|
| 1330 |
+
"learning_rate": 5.278532608695652e-06,
|
| 1331 |
+
"loss": 10.7913,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.6550522648083623,
|
| 1336 |
+
"grad_norm": 0.09419780969619751,
|
| 1337 |
+
"learning_rate": 5.244565217391306e-06,
|
| 1338 |
+
"loss": 10.7917,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.6637630662020906,
|
| 1343 |
+
"grad_norm": 0.0949205756187439,
|
| 1344 |
+
"learning_rate": 5.210597826086957e-06,
|
| 1345 |
+
"loss": 10.791,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.6724738675958188,
|
| 1350 |
+
"grad_norm": 0.09357789158821106,
|
| 1351 |
+
"learning_rate": 5.176630434782609e-06,
|
| 1352 |
+
"loss": 10.791,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.6811846689895469,
|
| 1357 |
+
"grad_norm": 0.0988670364022255,
|
| 1358 |
+
"learning_rate": 5.142663043478261e-06,
|
| 1359 |
+
"loss": 10.7917,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.6898954703832754,
|
| 1364 |
+
"grad_norm": 0.09996125102043152,
|
| 1365 |
+
"learning_rate": 5.108695652173914e-06,
|
| 1366 |
+
"loss": 10.7912,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.6986062717770034,
|
| 1371 |
+
"grad_norm": 0.09966638684272766,
|
| 1372 |
+
"learning_rate": 5.074728260869566e-06,
|
| 1373 |
+
"loss": 10.825,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.7073170731707317,
|
| 1378 |
+
"grad_norm": 0.09682171791791916,
|
| 1379 |
+
"learning_rate": 5.040760869565218e-06,
|
| 1380 |
+
"loss": 10.7906,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.71602787456446,
|
| 1385 |
+
"grad_norm": 0.09694927930831909,
|
| 1386 |
+
"learning_rate": 5.0067934782608696e-06,
|
| 1387 |
+
"loss": 10.7911,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.7247386759581882,
|
| 1392 |
+
"grad_norm": 0.09398549050092697,
|
| 1393 |
+
"learning_rate": 4.972826086956522e-06,
|
| 1394 |
+
"loss": 10.7909,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.7334494773519165,
|
| 1399 |
+
"grad_norm": 0.08475075662136078,
|
| 1400 |
+
"learning_rate": 4.938858695652175e-06,
|
| 1401 |
+
"loss": 10.7916,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.7421602787456445,
|
| 1406 |
+
"grad_norm": 0.09494160860776901,
|
| 1407 |
+
"learning_rate": 4.9048913043478265e-06,
|
| 1408 |
+
"loss": 10.7911,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.750871080139373,
|
| 1413 |
+
"grad_norm": 0.08939459174871445,
|
| 1414 |
+
"learning_rate": 4.870923913043479e-06,
|
| 1415 |
+
"loss": 10.7906,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.759581881533101,
|
| 1420 |
+
"grad_norm": 0.09646463394165039,
|
| 1421 |
+
"learning_rate": 4.836956521739131e-06,
|
| 1422 |
+
"loss": 10.7911,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.7682926829268293,
|
| 1427 |
+
"grad_norm": 0.09423504769802094,
|
| 1428 |
+
"learning_rate": 4.8029891304347835e-06,
|
| 1429 |
+
"loss": 10.7906,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.7770034843205575,
|
| 1434 |
+
"grad_norm": 0.09972324222326279,
|
| 1435 |
+
"learning_rate": 4.769021739130435e-06,
|
| 1436 |
+
"loss": 10.7914,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.7857142857142856,
|
| 1441 |
+
"grad_norm": 0.10339873284101486,
|
| 1442 |
+
"learning_rate": 4.735054347826088e-06,
|
| 1443 |
+
"loss": 10.7898,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.794425087108014,
|
| 1448 |
+
"grad_norm": 0.09187936782836914,
|
| 1449 |
+
"learning_rate": 4.7010869565217396e-06,
|
| 1450 |
+
"loss": 10.7915,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.8031358885017421,
|
| 1455 |
+
"grad_norm": 0.09024845063686371,
|
| 1456 |
+
"learning_rate": 4.667119565217391e-06,
|
| 1457 |
+
"loss": 10.7902,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.8118466898954704,
|
| 1462 |
+
"grad_norm": 0.08929532021284103,
|
| 1463 |
+
"learning_rate": 4.633152173913044e-06,
|
| 1464 |
+
"loss": 10.7905,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.8205574912891986,
|
| 1469 |
+
"grad_norm": 0.09708672016859055,
|
| 1470 |
+
"learning_rate": 4.599184782608696e-06,
|
| 1471 |
+
"loss": 10.7901,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.8292682926829267,
|
| 1476 |
+
"grad_norm": 0.10239408910274506,
|
| 1477 |
+
"learning_rate": 4.565217391304348e-06,
|
| 1478 |
+
"loss": 10.7896,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.8379790940766552,
|
| 1483 |
+
"grad_norm": 0.10148520767688751,
|
| 1484 |
+
"learning_rate": 4.53125e-06,
|
| 1485 |
+
"loss": 10.7906,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.8466898954703832,
|
| 1490 |
+
"grad_norm": 0.09579440951347351,
|
| 1491 |
+
"learning_rate": 4.497282608695653e-06,
|
| 1492 |
+
"loss": 10.7903,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.8554006968641115,
|
| 1497 |
+
"grad_norm": 0.09891035407781601,
|
| 1498 |
+
"learning_rate": 4.463315217391304e-06,
|
| 1499 |
+
"loss": 10.7904,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.8641114982578397,
|
| 1504 |
+
"grad_norm": 0.09411191940307617,
|
| 1505 |
+
"learning_rate": 4.429347826086957e-06,
|
| 1506 |
+
"loss": 10.7899,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.872822299651568,
|
| 1511 |
+
"grad_norm": 0.10932313650846481,
|
| 1512 |
+
"learning_rate": 4.395380434782609e-06,
|
| 1513 |
+
"loss": 10.7907,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.8815331010452963,
|
| 1518 |
+
"grad_norm": 0.08247391879558563,
|
| 1519 |
+
"learning_rate": 4.361413043478261e-06,
|
| 1520 |
+
"loss": 10.7902,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.8902439024390243,
|
| 1525 |
+
"grad_norm": 0.0989106297492981,
|
| 1526 |
+
"learning_rate": 4.327445652173913e-06,
|
| 1527 |
+
"loss": 10.79,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.8989547038327528,
|
| 1532 |
+
"grad_norm": 0.09022705256938934,
|
| 1533 |
+
"learning_rate": 4.293478260869566e-06,
|
| 1534 |
+
"loss": 10.7893,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.9076655052264808,
|
| 1539 |
+
"grad_norm": 0.09723115712404251,
|
| 1540 |
+
"learning_rate": 4.2595108695652174e-06,
|
| 1541 |
+
"loss": 10.7892,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.916376306620209,
|
| 1546 |
+
"grad_norm": 0.09660723060369492,
|
| 1547 |
+
"learning_rate": 4.22554347826087e-06,
|
| 1548 |
+
"loss": 10.7899,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.9250871080139373,
|
| 1553 |
+
"grad_norm": 0.09882649034261703,
|
| 1554 |
+
"learning_rate": 4.191576086956522e-06,
|
| 1555 |
+
"loss": 10.7896,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.9337979094076654,
|
| 1560 |
+
"grad_norm": 0.08623263239860535,
|
| 1561 |
+
"learning_rate": 4.157608695652174e-06,
|
| 1562 |
+
"loss": 10.7894,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.9425087108013939,
|
| 1567 |
+
"grad_norm": 0.09180337190628052,
|
| 1568 |
+
"learning_rate": 4.123641304347826e-06,
|
| 1569 |
+
"loss": 10.7896,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.951219512195122,
|
| 1574 |
+
"grad_norm": 0.09379716962575912,
|
| 1575 |
+
"learning_rate": 4.089673913043479e-06,
|
| 1576 |
+
"loss": 10.79,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.9599303135888502,
|
| 1581 |
+
"grad_norm": 0.09439699351787567,
|
| 1582 |
+
"learning_rate": 4.0557065217391305e-06,
|
| 1583 |
+
"loss": 10.79,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.9686411149825784,
|
| 1588 |
+
"grad_norm": 0.09595078229904175,
|
| 1589 |
+
"learning_rate": 4.021739130434783e-06,
|
| 1590 |
+
"loss": 10.7891,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.9773519163763065,
|
| 1595 |
+
"grad_norm": 0.09709908813238144,
|
| 1596 |
+
"learning_rate": 3.987771739130435e-06,
|
| 1597 |
+
"loss": 10.7899,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.986062717770035,
|
| 1602 |
+
"grad_norm": 0.0975862666964531,
|
| 1603 |
+
"learning_rate": 3.9538043478260875e-06,
|
| 1604 |
+
"loss": 10.7885,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.994773519163763,
|
| 1609 |
+
"grad_norm": 0.09871925413608551,
|
| 1610 |
+
"learning_rate": 3.919836956521739e-06,
|
| 1611 |
+
"loss": 10.7895,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.0034843205574915,
|
| 1616 |
+
"grad_norm": 0.09349826723337173,
|
| 1617 |
+
"learning_rate": 3.885869565217392e-06,
|
| 1618 |
+
"loss": 10.7883,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.0121951219512195,
|
| 1623 |
+
"grad_norm": 0.09288814663887024,
|
| 1624 |
+
"learning_rate": 3.8519021739130436e-06,
|
| 1625 |
+
"loss": 10.7887,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.0209059233449476,
|
| 1630 |
+
"grad_norm": 0.093532495200634,
|
| 1631 |
+
"learning_rate": 3.817934782608696e-06,
|
| 1632 |
+
"loss": 10.7892,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.029616724738676,
|
| 1637 |
+
"grad_norm": 0.10073115676641464,
|
| 1638 |
+
"learning_rate": 3.783967391304348e-06,
|
| 1639 |
+
"loss": 10.789,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.038327526132404,
|
| 1644 |
+
"grad_norm": 0.09286133199930191,
|
| 1645 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 1646 |
+
"loss": 10.7899,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.0470383275261326,
|
| 1651 |
+
"grad_norm": 0.08912254869937897,
|
| 1652 |
+
"learning_rate": 3.7160326086956527e-06,
|
| 1653 |
+
"loss": 10.7896,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.0557491289198606,
|
| 1658 |
+
"grad_norm": 0.09220988303422928,
|
| 1659 |
+
"learning_rate": 3.6820652173913044e-06,
|
| 1660 |
+
"loss": 10.7892,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.0644599303135887,
|
| 1665 |
+
"grad_norm": 67073.5859375,
|
| 1666 |
+
"learning_rate": 3.648097826086957e-06,
|
| 1667 |
+
"loss": 12.4148,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.073170731707317,
|
| 1672 |
+
"grad_norm": 0.09415777027606964,
|
| 1673 |
+
"learning_rate": 3.614130434782609e-06,
|
| 1674 |
+
"loss": 12.1967,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.081881533101045,
|
| 1679 |
+
"grad_norm": 0.09904878586530685,
|
| 1680 |
+
"learning_rate": 3.5801630434782614e-06,
|
| 1681 |
+
"loss": 10.7889,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.0905923344947737,
|
| 1686 |
+
"grad_norm": 0.09543213993310928,
|
| 1687 |
+
"learning_rate": 3.546195652173913e-06,
|
| 1688 |
+
"loss": 10.7888,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.0993031358885017,
|
| 1693 |
+
"grad_norm": 0.09227117896080017,
|
| 1694 |
+
"learning_rate": 3.5122282608695658e-06,
|
| 1695 |
+
"loss": 10.7894,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.10801393728223,
|
| 1700 |
+
"grad_norm": 0.09841565787792206,
|
| 1701 |
+
"learning_rate": 3.4782608695652175e-06,
|
| 1702 |
+
"loss": 10.7894,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.1167247386759582,
|
| 1707 |
+
"grad_norm": 0.093239426612854,
|
| 1708 |
+
"learning_rate": 3.44429347826087e-06,
|
| 1709 |
+
"loss": 10.7894,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.1254355400696863,
|
| 1714 |
+
"grad_norm": 0.09383922815322876,
|
| 1715 |
+
"learning_rate": 3.410326086956522e-06,
|
| 1716 |
+
"loss": 10.7882,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.1341463414634148,
|
| 1721 |
+
"grad_norm": 0.09073406457901001,
|
| 1722 |
+
"learning_rate": 3.3763586956521745e-06,
|
| 1723 |
+
"loss": 10.7897,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.142857142857143,
|
| 1728 |
+
"grad_norm": 0.09569002687931061,
|
| 1729 |
+
"learning_rate": 3.3423913043478262e-06,
|
| 1730 |
+
"loss": 10.7897,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.1515679442508713,
|
| 1735 |
+
"grad_norm": 0.0955529659986496,
|
| 1736 |
+
"learning_rate": 3.308423913043479e-06,
|
| 1737 |
+
"loss": 10.7883,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.1602787456445993,
|
| 1742 |
+
"grad_norm": 0.10046317428350449,
|
| 1743 |
+
"learning_rate": 3.2744565217391306e-06,
|
| 1744 |
+
"loss": 10.7895,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.1689895470383274,
|
| 1749 |
+
"grad_norm": 0.09671170264482498,
|
| 1750 |
+
"learning_rate": 3.240489130434783e-06,
|
| 1751 |
+
"loss": 10.789,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.177700348432056,
|
| 1756 |
+
"grad_norm": 0.09719261527061462,
|
| 1757 |
+
"learning_rate": 3.206521739130435e-06,
|
| 1758 |
+
"loss": 10.788,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.186411149825784,
|
| 1763 |
+
"grad_norm": 0.09898174554109573,
|
| 1764 |
+
"learning_rate": 3.172554347826087e-06,
|
| 1765 |
+
"loss": 10.788,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.1951219512195124,
|
| 1770 |
+
"grad_norm": 0.09254367649555206,
|
| 1771 |
+
"learning_rate": 3.1385869565217393e-06,
|
| 1772 |
+
"loss": 10.788,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.2038327526132404,
|
| 1777 |
+
"grad_norm": 0.10173038393259048,
|
| 1778 |
+
"learning_rate": 3.1046195652173915e-06,
|
| 1779 |
+
"loss": 10.7876,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.2125435540069684,
|
| 1784 |
+
"grad_norm": 0.0905982032418251,
|
| 1785 |
+
"learning_rate": 3.0706521739130436e-06,
|
| 1786 |
+
"loss": 10.7889,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.221254355400697,
|
| 1791 |
+
"grad_norm": 0.09394685924053192,
|
| 1792 |
+
"learning_rate": 3.036684782608696e-06,
|
| 1793 |
+
"loss": 10.7884,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.229965156794425,
|
| 1798 |
+
"grad_norm": 0.10268858075141907,
|
| 1799 |
+
"learning_rate": 3.002717391304348e-06,
|
| 1800 |
+
"loss": 10.801,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.2386759581881535,
|
| 1805 |
+
"grad_norm": 0.08987828344106674,
|
| 1806 |
+
"learning_rate": 2.96875e-06,
|
| 1807 |
+
"loss": 10.7892,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.2473867595818815,
|
| 1812 |
+
"grad_norm": 0.09386924654245377,
|
| 1813 |
+
"learning_rate": 2.9347826086956528e-06,
|
| 1814 |
+
"loss": 10.7879,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.2560975609756095,
|
| 1819 |
+
"grad_norm": 0.08885066956281662,
|
| 1820 |
+
"learning_rate": 2.9008152173913045e-06,
|
| 1821 |
+
"loss": 10.7887,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.264808362369338,
|
| 1826 |
+
"grad_norm": 0.09525702148675919,
|
| 1827 |
+
"learning_rate": 2.866847826086957e-06,
|
| 1828 |
+
"loss": 10.7885,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.273519163763066,
|
| 1833 |
+
"grad_norm": 0.09443964064121246,
|
| 1834 |
+
"learning_rate": 2.832880434782609e-06,
|
| 1835 |
+
"loss": 10.7887,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.2822299651567945,
|
| 1840 |
+
"grad_norm": 0.09282263368368149,
|
| 1841 |
+
"learning_rate": 2.7989130434782615e-06,
|
| 1842 |
+
"loss": 10.7889,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.2909407665505226,
|
| 1847 |
+
"grad_norm": 0.09663678705692291,
|
| 1848 |
+
"learning_rate": 2.7649456521739132e-06,
|
| 1849 |
+
"loss": 10.7885,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.2996515679442506,
|
| 1854 |
+
"grad_norm": 0.09053708612918854,
|
| 1855 |
+
"learning_rate": 2.730978260869566e-06,
|
| 1856 |
+
"loss": 10.7881,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.308362369337979,
|
| 1861 |
+
"grad_norm": 0.09515667706727982,
|
| 1862 |
+
"learning_rate": 2.6970108695652176e-06,
|
| 1863 |
+
"loss": 10.7882,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.317073170731707,
|
| 1868 |
+
"grad_norm": 0.09767191112041473,
|
| 1869 |
+
"learning_rate": 2.6630434782608698e-06,
|
| 1870 |
+
"loss": 10.7886,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.3257839721254356,
|
| 1875 |
+
"grad_norm": 0.09252949804067612,
|
| 1876 |
+
"learning_rate": 2.629076086956522e-06,
|
| 1877 |
+
"loss": 10.7883,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.3344947735191637,
|
| 1882 |
+
"grad_norm": 0.09851150959730148,
|
| 1883 |
+
"learning_rate": 2.595108695652174e-06,
|
| 1884 |
+
"loss": 10.7878,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.343205574912892,
|
| 1889 |
+
"grad_norm": 0.09727566689252853,
|
| 1890 |
+
"learning_rate": 2.5611413043478263e-06,
|
| 1891 |
+
"loss": 10.7894,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.35191637630662,
|
| 1896 |
+
"grad_norm": 0.09644806385040283,
|
| 1897 |
+
"learning_rate": 2.5271739130434785e-06,
|
| 1898 |
+
"loss": 10.7872,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.3606271777003482,
|
| 1903 |
+
"grad_norm": 0.09369703382253647,
|
| 1904 |
+
"learning_rate": 2.4932065217391306e-06,
|
| 1905 |
+
"loss": 10.788,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.3693379790940767,
|
| 1910 |
+
"grad_norm": 0.0991264134645462,
|
| 1911 |
+
"learning_rate": 2.459239130434783e-06,
|
| 1912 |
+
"loss": 10.788,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.3780487804878048,
|
| 1917 |
+
"grad_norm": 0.09739573299884796,
|
| 1918 |
+
"learning_rate": 2.425271739130435e-06,
|
| 1919 |
+
"loss": 10.7879,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.3867595818815333,
|
| 1924 |
+
"grad_norm": 0.10022178292274475,
|
| 1925 |
+
"learning_rate": 2.391304347826087e-06,
|
| 1926 |
+
"loss": 10.7881,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.3954703832752613,
|
| 1931 |
+
"grad_norm": 0.09441060572862625,
|
| 1932 |
+
"learning_rate": 2.3573369565217393e-06,
|
| 1933 |
+
"loss": 10.7879,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.40418118466899,
|
| 1938 |
+
"grad_norm": 0.09166783839464188,
|
| 1939 |
+
"learning_rate": 2.3233695652173915e-06,
|
| 1940 |
+
"loss": 10.7873,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.412891986062718,
|
| 1945 |
+
"grad_norm": 0.09163929522037506,
|
| 1946 |
+
"learning_rate": 2.2894021739130437e-06,
|
| 1947 |
+
"loss": 10.7878,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.421602787456446,
|
| 1952 |
+
"grad_norm": 0.09094561636447906,
|
| 1953 |
+
"learning_rate": 2.255434782608696e-06,
|
| 1954 |
+
"loss": 10.7885,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.4303135888501743,
|
| 1959 |
+
"grad_norm": 0.09401146322488785,
|
| 1960 |
+
"learning_rate": 2.221467391304348e-06,
|
| 1961 |
+
"loss": 10.7882,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.4390243902439024,
|
| 1966 |
+
"grad_norm": 0.08646171540021896,
|
| 1967 |
+
"learning_rate": 2.1875000000000002e-06,
|
| 1968 |
+
"loss": 10.7958,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.447735191637631,
|
| 1973 |
+
"grad_norm": 0.09187670797109604,
|
| 1974 |
+
"learning_rate": 2.1535326086956524e-06,
|
| 1975 |
+
"loss": 10.7879,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.456445993031359,
|
| 1980 |
+
"grad_norm": 0.09049776941537857,
|
| 1981 |
+
"learning_rate": 2.1195652173913046e-06,
|
| 1982 |
+
"loss": 10.7887,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.465156794425087,
|
| 1987 |
+
"grad_norm": 0.09695518016815186,
|
| 1988 |
+
"learning_rate": 2.0855978260869568e-06,
|
| 1989 |
+
"loss": 10.7878,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.4738675958188154,
|
| 1994 |
+
"grad_norm": 0.09298545867204666,
|
| 1995 |
+
"learning_rate": 2.051630434782609e-06,
|
| 1996 |
+
"loss": 10.7885,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.4825783972125435,
|
| 2001 |
+
"grad_norm": 0.09423473477363586,
|
| 2002 |
+
"learning_rate": 2.017663043478261e-06,
|
| 2003 |
+
"loss": 10.7879,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.491289198606272,
|
| 2008 |
+
"grad_norm": 0.0882716029882431,
|
| 2009 |
+
"learning_rate": 1.9836956521739133e-06,
|
| 2010 |
+
"loss": 10.7877,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.5,
|
| 2015 |
+
"grad_norm": 0.09123384207487106,
|
| 2016 |
+
"learning_rate": 1.9497282608695655e-06,
|
| 2017 |
+
"loss": 10.788,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.508710801393728,
|
| 2022 |
+
"grad_norm": 0.09263134002685547,
|
| 2023 |
+
"learning_rate": 1.9157608695652176e-06,
|
| 2024 |
+
"loss": 10.788,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.5174216027874565,
|
| 2029 |
+
"grad_norm": 0.09086143225431442,
|
| 2030 |
+
"learning_rate": 1.8817934782608696e-06,
|
| 2031 |
+
"loss": 10.7876,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.5261324041811846,
|
| 2036 |
+
"grad_norm": 0.10285180062055588,
|
| 2037 |
+
"learning_rate": 1.8478260869565218e-06,
|
| 2038 |
+
"loss": 10.7872,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.534843205574913,
|
| 2043 |
+
"grad_norm": 0.09717818349599838,
|
| 2044 |
+
"learning_rate": 1.813858695652174e-06,
|
| 2045 |
+
"loss": 10.7878,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.543554006968641,
|
| 2050 |
+
"grad_norm": 0.09603049606084824,
|
| 2051 |
+
"learning_rate": 1.7798913043478264e-06,
|
| 2052 |
+
"loss": 10.7877,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.552264808362369,
|
| 2057 |
+
"grad_norm": 0.09593737125396729,
|
| 2058 |
+
"learning_rate": 1.7459239130434785e-06,
|
| 2059 |
+
"loss": 10.7865,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.5609756097560976,
|
| 2064 |
+
"grad_norm": 0.0944036915898323,
|
| 2065 |
+
"learning_rate": 1.7119565217391307e-06,
|
| 2066 |
+
"loss": 10.7876,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.5696864111498257,
|
| 2071 |
+
"grad_norm": 0.09109403192996979,
|
| 2072 |
+
"learning_rate": 1.6779891304347829e-06,
|
| 2073 |
+
"loss": 10.7866,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.578397212543554,
|
| 2078 |
+
"grad_norm": 0.09073559939861298,
|
| 2079 |
+
"learning_rate": 1.644021739130435e-06,
|
| 2080 |
+
"loss": 10.7882,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.587108013937282,
|
| 2085 |
+
"grad_norm": 0.09691879898309708,
|
| 2086 |
+
"learning_rate": 1.6100543478260872e-06,
|
| 2087 |
+
"loss": 10.7886,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.59581881533101,
|
| 2092 |
+
"grad_norm": 0.09810439497232437,
|
| 2093 |
+
"learning_rate": 1.5760869565217394e-06,
|
| 2094 |
+
"loss": 10.7873,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.6045296167247387,
|
| 2099 |
+
"grad_norm": 0.09767334908246994,
|
| 2100 |
+
"learning_rate": 1.5421195652173914e-06,
|
| 2101 |
+
"loss": 10.7872,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.6132404181184667,
|
| 2106 |
+
"grad_norm": 0.09681523591279984,
|
| 2107 |
+
"learning_rate": 1.5081521739130436e-06,
|
| 2108 |
+
"loss": 10.7872,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.6219512195121952,
|
| 2113 |
+
"grad_norm": 0.10315357148647308,
|
| 2114 |
+
"learning_rate": 1.4741847826086957e-06,
|
| 2115 |
+
"loss": 10.7877,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.6306620209059233,
|
| 2120 |
+
"grad_norm": 0.09027998894453049,
|
| 2121 |
+
"learning_rate": 1.440217391304348e-06,
|
| 2122 |
+
"loss": 10.794,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.6393728222996513,
|
| 2127 |
+
"grad_norm": 0.09546509385108948,
|
| 2128 |
+
"learning_rate": 1.40625e-06,
|
| 2129 |
+
"loss": 10.7875,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.64808362369338,
|
| 2134 |
+
"grad_norm": 0.09259118884801865,
|
| 2135 |
+
"learning_rate": 1.3722826086956523e-06,
|
| 2136 |
+
"loss": 10.787,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.6567944250871083,
|
| 2141 |
+
"grad_norm": 0.09348276257514954,
|
| 2142 |
+
"learning_rate": 1.3383152173913044e-06,
|
| 2143 |
+
"loss": 10.7886,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.6655052264808363,
|
| 2148 |
+
"grad_norm": 0.09209384769201279,
|
| 2149 |
+
"learning_rate": 1.3043478260869566e-06,
|
| 2150 |
+
"loss": 10.7878,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.6742160278745644,
|
| 2155 |
+
"grad_norm": 0.09436801075935364,
|
| 2156 |
+
"learning_rate": 1.2703804347826088e-06,
|
| 2157 |
+
"loss": 10.7879,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.682926829268293,
|
| 2162 |
+
"grad_norm": 0.09730960428714752,
|
| 2163 |
+
"learning_rate": 1.236413043478261e-06,
|
| 2164 |
+
"loss": 10.7882,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.691637630662021,
|
| 2169 |
+
"grad_norm": 0.09457021951675415,
|
| 2170 |
+
"learning_rate": 1.2024456521739131e-06,
|
| 2171 |
+
"loss": 10.7878,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.7003484320557494,
|
| 2176 |
+
"grad_norm": 0.09415573626756668,
|
| 2177 |
+
"learning_rate": 1.1684782608695653e-06,
|
| 2178 |
+
"loss": 10.7877,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.7090592334494774,
|
| 2183 |
+
"grad_norm": 0.09336414933204651,
|
| 2184 |
+
"learning_rate": 1.1345108695652175e-06,
|
| 2185 |
+
"loss": 10.7863,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.7177700348432055,
|
| 2190 |
+
"grad_norm": 0.09788176417350769,
|
| 2191 |
+
"learning_rate": 1.1005434782608697e-06,
|
| 2192 |
+
"loss": 10.7866,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.726480836236934,
|
| 2197 |
+
"grad_norm": 0.10606147348880768,
|
| 2198 |
+
"learning_rate": 1.0665760869565219e-06,
|
| 2199 |
+
"loss": 10.788,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.735191637630662,
|
| 2204 |
+
"grad_norm": 0.09581784158945084,
|
| 2205 |
+
"learning_rate": 1.032608695652174e-06,
|
| 2206 |
+
"loss": 10.7865,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.7439024390243905,
|
| 2211 |
+
"grad_norm": 0.09258154779672623,
|
| 2212 |
+
"learning_rate": 9.986413043478262e-07,
|
| 2213 |
+
"loss": 10.7874,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.7526132404181185,
|
| 2218 |
+
"grad_norm": 0.09044299274682999,
|
| 2219 |
+
"learning_rate": 9.646739130434784e-07,
|
| 2220 |
+
"loss": 10.7878,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.7613240418118465,
|
| 2225 |
+
"grad_norm": 0.09484916925430298,
|
| 2226 |
+
"learning_rate": 9.307065217391305e-07,
|
| 2227 |
+
"loss": 10.787,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.770034843205575,
|
| 2232 |
+
"grad_norm": 0.0974380299448967,
|
| 2233 |
+
"learning_rate": 8.967391304347826e-07,
|
| 2234 |
+
"loss": 10.7871,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.778745644599303,
|
| 2239 |
+
"grad_norm": 0.093409463763237,
|
| 2240 |
+
"learning_rate": 8.627717391304348e-07,
|
| 2241 |
+
"loss": 10.7879,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.7874564459930316,
|
| 2246 |
+
"grad_norm": 0.09446702152490616,
|
| 2247 |
+
"learning_rate": 8.28804347826087e-07,
|
| 2248 |
+
"loss": 10.7865,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.7961672473867596,
|
| 2253 |
+
"grad_norm": 0.10579903423786163,
|
| 2254 |
+
"learning_rate": 7.948369565217393e-07,
|
| 2255 |
+
"loss": 10.7874,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.8048780487804876,
|
| 2260 |
+
"grad_norm": 0.09646815061569214,
|
| 2261 |
+
"learning_rate": 7.608695652173914e-07,
|
| 2262 |
+
"loss": 10.787,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.813588850174216,
|
| 2267 |
+
"grad_norm": 0.09360378235578537,
|
| 2268 |
+
"learning_rate": 7.269021739130436e-07,
|
| 2269 |
+
"loss": 10.7865,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.822299651567944,
|
| 2274 |
+
"grad_norm": 0.09116079658269882,
|
| 2275 |
+
"learning_rate": 6.929347826086957e-07,
|
| 2276 |
+
"loss": 10.7874,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.8310104529616726,
|
| 2281 |
+
"grad_norm": 0.09042556583881378,
|
| 2282 |
+
"learning_rate": 6.589673913043479e-07,
|
| 2283 |
+
"loss": 10.7871,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.8397212543554007,
|
| 2288 |
+
"grad_norm": 0.09959082305431366,
|
| 2289 |
+
"learning_rate": 6.25e-07,
|
| 2290 |
+
"loss": 10.7863,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.8484320557491287,
|
| 2295 |
+
"grad_norm": 0.09715747833251953,
|
| 2296 |
+
"learning_rate": 5.910326086956522e-07,
|
| 2297 |
+
"loss": 10.7879,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.857142857142857,
|
| 2302 |
+
"grad_norm": 0.09984558820724487,
|
| 2303 |
+
"learning_rate": 5.570652173913044e-07,
|
| 2304 |
+
"loss": 10.7865,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.8658536585365852,
|
| 2309 |
+
"grad_norm": 0.09477663040161133,
|
| 2310 |
+
"learning_rate": 5.230978260869566e-07,
|
| 2311 |
+
"loss": 10.7873,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.8745644599303137,
|
| 2316 |
+
"grad_norm": 0.09480907022953033,
|
| 2317 |
+
"learning_rate": 4.891304347826088e-07,
|
| 2318 |
+
"loss": 10.7874,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.8832752613240418,
|
| 2323 |
+
"grad_norm": 0.09444177150726318,
|
| 2324 |
+
"learning_rate": 4.5516304347826094e-07,
|
| 2325 |
+
"loss": 10.787,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.89198606271777,
|
| 2330 |
+
"grad_norm": 0.09742891043424606,
|
| 2331 |
+
"learning_rate": 4.2119565217391306e-07,
|
| 2332 |
+
"loss": 10.7869,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.9006968641114983,
|
| 2337 |
+
"grad_norm": 0.09473706036806107,
|
| 2338 |
+
"learning_rate": 3.8722826086956524e-07,
|
| 2339 |
+
"loss": 10.7873,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.9094076655052263,
|
| 2344 |
+
"grad_norm": 0.09218768030405045,
|
| 2345 |
+
"learning_rate": 3.532608695652174e-07,
|
| 2346 |
+
"loss": 10.7871,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.918118466898955,
|
| 2351 |
+
"grad_norm": 0.09080421179533005,
|
| 2352 |
+
"learning_rate": 3.1929347826086964e-07,
|
| 2353 |
+
"loss": 10.7868,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.926829268292683,
|
| 2358 |
+
"grad_norm": 0.09067242592573166,
|
| 2359 |
+
"learning_rate": 2.8532608695652177e-07,
|
| 2360 |
+
"loss": 10.787,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.935540069686411,
|
| 2365 |
+
"grad_norm": 0.09306979924440384,
|
| 2366 |
+
"learning_rate": 2.5135869565217394e-07,
|
| 2367 |
+
"loss": 10.7869,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.9442508710801394,
|
| 2372 |
+
"grad_norm": 0.09051218628883362,
|
| 2373 |
+
"learning_rate": 2.173913043478261e-07,
|
| 2374 |
+
"loss": 10.7861,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.952961672473868,
|
| 2379 |
+
"grad_norm": 0.09443790465593338,
|
| 2380 |
+
"learning_rate": 1.834239130434783e-07,
|
| 2381 |
+
"loss": 10.7871,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.961672473867596,
|
| 2386 |
+
"grad_norm": 0.09303121268749237,
|
| 2387 |
+
"learning_rate": 1.4945652173913045e-07,
|
| 2388 |
+
"loss": 10.7871,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.970383275261324,
|
| 2393 |
+
"grad_norm": 0.095026396214962,
|
| 2394 |
+
"learning_rate": 1.1548913043478261e-07,
|
| 2395 |
+
"loss": 10.7878,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.979094076655052,
|
| 2400 |
+
"grad_norm": 0.09600502252578735,
|
| 2401 |
+
"learning_rate": 8.152173913043479e-08,
|
| 2402 |
+
"loss": 10.7873,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.9878048780487805,
|
| 2407 |
+
"grad_norm": 0.09918372333049774,
|
| 2408 |
+
"learning_rate": 4.7554347826086966e-08,
|
| 2409 |
+
"loss": 10.7882,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.996515679442509,
|
| 2414 |
+
"grad_norm": 0.10064807534217834,
|
| 2415 |
+
"learning_rate": 1.3586956521739131e-08,
|
| 2416 |
+
"loss": 10.7876,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
}
|
| 2419 |
+
],
|
| 2420 |
+
"logging_steps": 10,
|
| 2421 |
+
"max_steps": 3444,
|
| 2422 |
+
"num_input_tokens_seen": 0,
|
| 2423 |
+
"num_train_epochs": 3,
|
| 2424 |
+
"save_steps": 500,
|
| 2425 |
+
"stateful_callbacks": {
|
| 2426 |
+
"TrainerControl": {
|
| 2427 |
+
"args": {
|
| 2428 |
+
"should_epoch_stop": false,
|
| 2429 |
+
"should_evaluate": false,
|
| 2430 |
+
"should_log": false,
|
| 2431 |
+
"should_save": true,
|
| 2432 |
+
"should_training_stop": true
|
| 2433 |
+
},
|
| 2434 |
+
"attributes": {}
|
| 2435 |
+
}
|
| 2436 |
+
},
|
| 2437 |
+
"total_flos": 1697588517943296.0,
|
| 2438 |
+
"train_batch_size": 32,
|
| 2439 |
+
"trial_name": null,
|
| 2440 |
+
"trial_params": null
|
| 2441 |
+
}
|
checkpoint-3444/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90c67833faf31d4402ee0893264f8627505d9d95e6510f92d119f2d25dc10219
|
| 3 |
+
size 5176
|
checkpoint-3444/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4592/config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TraXLMistralForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"dropout": 0.1,
|
| 6 |
+
"dynamic_routing": true,
|
| 7 |
+
"ff_expansion_factor": 4,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"id2label": {
|
| 10 |
+
"0": "LABEL_0",
|
| 11 |
+
"1": "LABEL_1",
|
| 12 |
+
"2": "LABEL_2",
|
| 13 |
+
"3": "LABEL_3",
|
| 14 |
+
"4": "LABEL_4"
|
| 15 |
+
},
|
| 16 |
+
"is_decoder": true,
|
| 17 |
+
"label2id": {
|
| 18 |
+
"LABEL_0": 0,
|
| 19 |
+
"LABEL_1": 1,
|
| 20 |
+
"LABEL_2": 2,
|
| 21 |
+
"LABEL_3": 3,
|
| 22 |
+
"LABEL_4": 4
|
| 23 |
+
},
|
| 24 |
+
"max_computation_steps": 5,
|
| 25 |
+
"max_len": 256,
|
| 26 |
+
"memory_size": 256,
|
| 27 |
+
"model_type": "TraXLMistral",
|
| 28 |
+
"n_head": 4,
|
| 29 |
+
"n_layer": 4,
|
| 30 |
+
"rnn_units": 128,
|
| 31 |
+
"sparse_attention": true,
|
| 32 |
+
"torch_dtype": "float32",
|
| 33 |
+
"transformers_version": "4.44.2",
|
| 34 |
+
"vocab_size": 50257
|
| 35 |
+
}
|
checkpoint-4592/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4592/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04c4fe7926f220f2bebec5a1eb077ffc03af6823c870616135f38480a2ee6fb3
|
| 3 |
+
size 194522976
|
checkpoint-4592/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f72398701fdbc4f6ad6394e48c1d9cc6d0d3d7090a154017b0853a0a15be446f
|
| 3 |
+
size 369911866
|
checkpoint-4592/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12366c4414f72b4ab644cfe18f3a52985826e1e581f3bedf0b08153a7a761ef2
|
| 3 |
+
size 14244
|
checkpoint-4592/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46e1cc70b7afaba1fae7528a28d123a0254fb4e0c8df2d75c0d3f9953a4a6559
|
| 3 |
+
size 1064
|
checkpoint-4592/special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|endoftext|>",
|
| 3 |
+
"eos_token": "<|endoftext|>",
|
| 4 |
+
"pad_token": "<|endoftext|>",
|
| 5 |
+
"unk_token": "<|endoftext|>"
|
| 6 |
+
}
|
checkpoint-4592/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4592/tokenizer_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"bos_token": "<|endoftext|>",
|
| 14 |
+
"clean_up_tokenization_spaces": true,
|
| 15 |
+
"eos_token": "<|endoftext|>",
|
| 16 |
+
"model_max_length": 1024,
|
| 17 |
+
"pad_token": "<|endoftext|>",
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>"
|
| 20 |
+
}
|
checkpoint-4592/trainer_state.json
ADDED
|
@@ -0,0 +1,3246 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4592,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008710801393728223,
|
| 13 |
+
"grad_norm": Infinity,
|
| 14 |
+
"learning_rate": 3.71949339942388e-07,
|
| 15 |
+
"loss": 23.9832,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.017421602787456445,
|
| 20 |
+
"grad_norm": 759.6871337890625,
|
| 21 |
+
"learning_rate": 1.2088353548127608e-06,
|
| 22 |
+
"loss": 68.5664,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02613240418118467,
|
| 27 |
+
"grad_norm": 1186.6964111328125,
|
| 28 |
+
"learning_rate": 2.045721369683134e-06,
|
| 29 |
+
"loss": 113.461,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03484320557491289,
|
| 34 |
+
"grad_norm": NaN,
|
| 35 |
+
"learning_rate": 2.045721369683134e-06,
|
| 36 |
+
"loss": 0.0,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04355400696864112,
|
| 41 |
+
"grad_norm": NaN,
|
| 42 |
+
"learning_rate": 2.045721369683134e-06,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05226480836236934,
|
| 48 |
+
"grad_norm": NaN,
|
| 49 |
+
"learning_rate": 2.045721369683134e-06,
|
| 50 |
+
"loss": 0.0,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06097560975609756,
|
| 55 |
+
"grad_norm": NaN,
|
| 56 |
+
"learning_rate": 2.045721369683134e-06,
|
| 57 |
+
"loss": 0.0,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06968641114982578,
|
| 62 |
+
"grad_norm": NaN,
|
| 63 |
+
"learning_rate": 2.045721369683134e-06,
|
| 64 |
+
"loss": 0.0,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.078397212543554,
|
| 69 |
+
"grad_norm": NaN,
|
| 70 |
+
"learning_rate": 2.045721369683134e-06,
|
| 71 |
+
"loss": 0.0,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08710801393728224,
|
| 76 |
+
"grad_norm": NaN,
|
| 77 |
+
"learning_rate": 2.045721369683134e-06,
|
| 78 |
+
"loss": 0.0,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09581881533101046,
|
| 83 |
+
"grad_norm": NaN,
|
| 84 |
+
"learning_rate": 2.045721369683134e-06,
|
| 85 |
+
"loss": 0.0,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10452961672473868,
|
| 90 |
+
"grad_norm": NaN,
|
| 91 |
+
"learning_rate": 2.045721369683134e-06,
|
| 92 |
+
"loss": 0.0,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1132404181184669,
|
| 97 |
+
"grad_norm": NaN,
|
| 98 |
+
"learning_rate": 2.045721369683134e-06,
|
| 99 |
+
"loss": 0.0,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.12195121951219512,
|
| 104 |
+
"grad_norm": NaN,
|
| 105 |
+
"learning_rate": 2.045721369683134e-06,
|
| 106 |
+
"loss": 0.0,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.13066202090592335,
|
| 111 |
+
"grad_norm": NaN,
|
| 112 |
+
"learning_rate": 2.045721369683134e-06,
|
| 113 |
+
"loss": 0.0,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13937282229965156,
|
| 118 |
+
"grad_norm": NaN,
|
| 119 |
+
"learning_rate": 2.045721369683134e-06,
|
| 120 |
+
"loss": 0.0,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1480836236933798,
|
| 125 |
+
"grad_norm": NaN,
|
| 126 |
+
"learning_rate": 2.045721369683134e-06,
|
| 127 |
+
"loss": 0.0,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.156794425087108,
|
| 132 |
+
"grad_norm": NaN,
|
| 133 |
+
"learning_rate": 2.045721369683134e-06,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16550522648083624,
|
| 139 |
+
"grad_norm": NaN,
|
| 140 |
+
"learning_rate": 2.045721369683134e-06,
|
| 141 |
+
"loss": 0.0,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.17421602787456447,
|
| 146 |
+
"grad_norm": NaN,
|
| 147 |
+
"learning_rate": 2.045721369683134e-06,
|
| 148 |
+
"loss": 0.0,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.18292682926829268,
|
| 153 |
+
"grad_norm": NaN,
|
| 154 |
+
"learning_rate": 2.045721369683134e-06,
|
| 155 |
+
"loss": 0.0,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1916376306620209,
|
| 160 |
+
"grad_norm": NaN,
|
| 161 |
+
"learning_rate": 2.045721369683134e-06,
|
| 162 |
+
"loss": 0.0,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.20034843205574912,
|
| 167 |
+
"grad_norm": NaN,
|
| 168 |
+
"learning_rate": 2.045721369683134e-06,
|
| 169 |
+
"loss": 0.0,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20905923344947736,
|
| 174 |
+
"grad_norm": NaN,
|
| 175 |
+
"learning_rate": 2.045721369683134e-06,
|
| 176 |
+
"loss": 0.0,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21777003484320556,
|
| 181 |
+
"grad_norm": NaN,
|
| 182 |
+
"learning_rate": 2.045721369683134e-06,
|
| 183 |
+
"loss": 0.0,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2264808362369338,
|
| 188 |
+
"grad_norm": NaN,
|
| 189 |
+
"learning_rate": 2.045721369683134e-06,
|
| 190 |
+
"loss": 0.0,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.23519163763066203,
|
| 195 |
+
"grad_norm": NaN,
|
| 196 |
+
"learning_rate": 2.045721369683134e-06,
|
| 197 |
+
"loss": 0.0,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.24390243902439024,
|
| 202 |
+
"grad_norm": NaN,
|
| 203 |
+
"learning_rate": 2.045721369683134e-06,
|
| 204 |
+
"loss": 0.0,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.25261324041811845,
|
| 209 |
+
"grad_norm": NaN,
|
| 210 |
+
"learning_rate": 2.045721369683134e-06,
|
| 211 |
+
"loss": 0.0,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2613240418118467,
|
| 216 |
+
"grad_norm": NaN,
|
| 217 |
+
"learning_rate": 2.045721369683134e-06,
|
| 218 |
+
"loss": 0.0,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2700348432055749,
|
| 223 |
+
"grad_norm": NaN,
|
| 224 |
+
"learning_rate": 2.045721369683134e-06,
|
| 225 |
+
"loss": 0.0,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2787456445993031,
|
| 230 |
+
"grad_norm": NaN,
|
| 231 |
+
"learning_rate": 2.045721369683134e-06,
|
| 232 |
+
"loss": 0.0,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2874564459930314,
|
| 237 |
+
"grad_norm": NaN,
|
| 238 |
+
"learning_rate": 2.045721369683134e-06,
|
| 239 |
+
"loss": 0.0,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2961672473867596,
|
| 244 |
+
"grad_norm": NaN,
|
| 245 |
+
"learning_rate": 2.045721369683134e-06,
|
| 246 |
+
"loss": 0.0,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3048780487804878,
|
| 251 |
+
"grad_norm": NaN,
|
| 252 |
+
"learning_rate": 2.045721369683134e-06,
|
| 253 |
+
"loss": 0.0,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.313588850174216,
|
| 258 |
+
"grad_norm": NaN,
|
| 259 |
+
"learning_rate": 2.045721369683134e-06,
|
| 260 |
+
"loss": 0.0,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.32229965156794427,
|
| 265 |
+
"grad_norm": NaN,
|
| 266 |
+
"learning_rate": 2.045721369683134e-06,
|
| 267 |
+
"loss": 0.0,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3310104529616725,
|
| 272 |
+
"grad_norm": NaN,
|
| 273 |
+
"learning_rate": 2.045721369683134e-06,
|
| 274 |
+
"loss": 0.0,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3397212543554007,
|
| 279 |
+
"grad_norm": NaN,
|
| 280 |
+
"learning_rate": 2.045721369683134e-06,
|
| 281 |
+
"loss": 0.0,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.34843205574912894,
|
| 286 |
+
"grad_norm": NaN,
|
| 287 |
+
"learning_rate": 2.045721369683134e-06,
|
| 288 |
+
"loss": 0.0,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.35714285714285715,
|
| 293 |
+
"grad_norm": NaN,
|
| 294 |
+
"learning_rate": 2.045721369683134e-06,
|
| 295 |
+
"loss": 0.0,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.36585365853658536,
|
| 300 |
+
"grad_norm": NaN,
|
| 301 |
+
"learning_rate": 2.045721369683134e-06,
|
| 302 |
+
"loss": 0.0,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.37456445993031356,
|
| 307 |
+
"grad_norm": NaN,
|
| 308 |
+
"learning_rate": 2.045721369683134e-06,
|
| 309 |
+
"loss": 0.0,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3832752613240418,
|
| 314 |
+
"grad_norm": NaN,
|
| 315 |
+
"learning_rate": 2.045721369683134e-06,
|
| 316 |
+
"loss": 0.0,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.39198606271777003,
|
| 321 |
+
"grad_norm": NaN,
|
| 322 |
+
"learning_rate": 2.045721369683134e-06,
|
| 323 |
+
"loss": 0.0,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.40069686411149824,
|
| 328 |
+
"grad_norm": NaN,
|
| 329 |
+
"learning_rate": 2.045721369683134e-06,
|
| 330 |
+
"loss": 0.0,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.4094076655052265,
|
| 335 |
+
"grad_norm": NaN,
|
| 336 |
+
"learning_rate": 2.045721369683134e-06,
|
| 337 |
+
"loss": 0.0,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4181184668989547,
|
| 342 |
+
"grad_norm": NaN,
|
| 343 |
+
"learning_rate": 2.045721369683134e-06,
|
| 344 |
+
"loss": 0.0,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4268292682926829,
|
| 349 |
+
"grad_norm": NaN,
|
| 350 |
+
"learning_rate": 2.045721369683134e-06,
|
| 351 |
+
"loss": 0.0,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4355400696864111,
|
| 356 |
+
"grad_norm": NaN,
|
| 357 |
+
"learning_rate": 2.045721369683134e-06,
|
| 358 |
+
"loss": 0.0,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4442508710801394,
|
| 363 |
+
"grad_norm": NaN,
|
| 364 |
+
"learning_rate": 2.045721369683134e-06,
|
| 365 |
+
"loss": 0.0,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4529616724738676,
|
| 370 |
+
"grad_norm": NaN,
|
| 371 |
+
"learning_rate": 2.045721369683134e-06,
|
| 372 |
+
"loss": 0.0,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4616724738675958,
|
| 377 |
+
"grad_norm": NaN,
|
| 378 |
+
"learning_rate": 2.045721369683134e-06,
|
| 379 |
+
"loss": 0.0,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.47038327526132406,
|
| 384 |
+
"grad_norm": NaN,
|
| 385 |
+
"learning_rate": 2.045721369683134e-06,
|
| 386 |
+
"loss": 0.0,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.47909407665505227,
|
| 391 |
+
"grad_norm": NaN,
|
| 392 |
+
"learning_rate": 2.045721369683134e-06,
|
| 393 |
+
"loss": 0.0,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4878048780487805,
|
| 398 |
+
"grad_norm": NaN,
|
| 399 |
+
"learning_rate": 2.045721369683134e-06,
|
| 400 |
+
"loss": 0.0,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4965156794425087,
|
| 405 |
+
"grad_norm": NaN,
|
| 406 |
+
"learning_rate": 2.045721369683134e-06,
|
| 407 |
+
"loss": 0.0,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5052264808362369,
|
| 412 |
+
"grad_norm": NaN,
|
| 413 |
+
"learning_rate": 2.045721369683134e-06,
|
| 414 |
+
"loss": 0.0,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5139372822299652,
|
| 419 |
+
"grad_norm": NaN,
|
| 420 |
+
"learning_rate": 2.045721369683134e-06,
|
| 421 |
+
"loss": 0.0,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5226480836236934,
|
| 426 |
+
"grad_norm": NaN,
|
| 427 |
+
"learning_rate": 2.045721369683134e-06,
|
| 428 |
+
"loss": 0.0,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5313588850174216,
|
| 433 |
+
"grad_norm": NaN,
|
| 434 |
+
"learning_rate": 2.045721369683134e-06,
|
| 435 |
+
"loss": 0.0,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5400696864111498,
|
| 440 |
+
"grad_norm": NaN,
|
| 441 |
+
"learning_rate": 2.045721369683134e-06,
|
| 442 |
+
"loss": 0.0,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5487804878048781,
|
| 447 |
+
"grad_norm": NaN,
|
| 448 |
+
"learning_rate": 2.045721369683134e-06,
|
| 449 |
+
"loss": 0.0,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5574912891986062,
|
| 454 |
+
"grad_norm": NaN,
|
| 455 |
+
"learning_rate": 2.045721369683134e-06,
|
| 456 |
+
"loss": 0.0,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5662020905923345,
|
| 461 |
+
"grad_norm": NaN,
|
| 462 |
+
"learning_rate": 2.045721369683134e-06,
|
| 463 |
+
"loss": 0.0,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5749128919860628,
|
| 468 |
+
"grad_norm": NaN,
|
| 469 |
+
"learning_rate": 2.045721369683134e-06,
|
| 470 |
+
"loss": 0.0,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5836236933797909,
|
| 475 |
+
"grad_norm": NaN,
|
| 476 |
+
"learning_rate": 2.045721369683134e-06,
|
| 477 |
+
"loss": 0.0,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5923344947735192,
|
| 482 |
+
"grad_norm": NaN,
|
| 483 |
+
"learning_rate": 2.045721369683134e-06,
|
| 484 |
+
"loss": 0.0,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6010452961672473,
|
| 489 |
+
"grad_norm": NaN,
|
| 490 |
+
"learning_rate": 2.045721369683134e-06,
|
| 491 |
+
"loss": 0.0,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6097560975609756,
|
| 496 |
+
"grad_norm": NaN,
|
| 497 |
+
"learning_rate": 2.045721369683134e-06,
|
| 498 |
+
"loss": 0.0,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6184668989547039,
|
| 503 |
+
"grad_norm": NaN,
|
| 504 |
+
"learning_rate": 2.045721369683134e-06,
|
| 505 |
+
"loss": 0.0,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.627177700348432,
|
| 510 |
+
"grad_norm": NaN,
|
| 511 |
+
"learning_rate": 2.045721369683134e-06,
|
| 512 |
+
"loss": 0.0,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6358885017421603,
|
| 517 |
+
"grad_norm": NaN,
|
| 518 |
+
"learning_rate": 2.045721369683134e-06,
|
| 519 |
+
"loss": 0.0,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6445993031358885,
|
| 524 |
+
"grad_norm": NaN,
|
| 525 |
+
"learning_rate": 2.045721369683134e-06,
|
| 526 |
+
"loss": 0.0,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6533101045296167,
|
| 531 |
+
"grad_norm": NaN,
|
| 532 |
+
"learning_rate": 2.045721369683134e-06,
|
| 533 |
+
"loss": 0.0,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.662020905923345,
|
| 538 |
+
"grad_norm": NaN,
|
| 539 |
+
"learning_rate": 2.045721369683134e-06,
|
| 540 |
+
"loss": 0.0,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6707317073170732,
|
| 545 |
+
"grad_norm": NaN,
|
| 546 |
+
"learning_rate": 2.045721369683134e-06,
|
| 547 |
+
"loss": 0.0,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6794425087108014,
|
| 552 |
+
"grad_norm": NaN,
|
| 553 |
+
"learning_rate": 2.045721369683134e-06,
|
| 554 |
+
"loss": 0.0,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6881533101045296,
|
| 559 |
+
"grad_norm": NaN,
|
| 560 |
+
"learning_rate": 2.045721369683134e-06,
|
| 561 |
+
"loss": 0.0,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6968641114982579,
|
| 566 |
+
"grad_norm": NaN,
|
| 567 |
+
"learning_rate": 2.045721369683134e-06,
|
| 568 |
+
"loss": 0.0,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.705574912891986,
|
| 573 |
+
"grad_norm": NaN,
|
| 574 |
+
"learning_rate": 2.045721369683134e-06,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7142857142857143,
|
| 580 |
+
"grad_norm": NaN,
|
| 581 |
+
"learning_rate": 2.045721369683134e-06,
|
| 582 |
+
"loss": 0.0,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.7229965156794426,
|
| 587 |
+
"grad_norm": NaN,
|
| 588 |
+
"learning_rate": 2.045721369683134e-06,
|
| 589 |
+
"loss": 0.0,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.7317073170731707,
|
| 594 |
+
"grad_norm": NaN,
|
| 595 |
+
"learning_rate": 2.045721369683134e-06,
|
| 596 |
+
"loss": 0.0,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.740418118466899,
|
| 601 |
+
"grad_norm": NaN,
|
| 602 |
+
"learning_rate": 2.045721369683134e-06,
|
| 603 |
+
"loss": 0.0,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.7491289198606271,
|
| 608 |
+
"grad_norm": NaN,
|
| 609 |
+
"learning_rate": 2.045721369683134e-06,
|
| 610 |
+
"loss": 0.0,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7578397212543554,
|
| 615 |
+
"grad_norm": NaN,
|
| 616 |
+
"learning_rate": 2.045721369683134e-06,
|
| 617 |
+
"loss": 0.0,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7665505226480837,
|
| 622 |
+
"grad_norm": NaN,
|
| 623 |
+
"learning_rate": 2.045721369683134e-06,
|
| 624 |
+
"loss": 0.0,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7752613240418118,
|
| 629 |
+
"grad_norm": NaN,
|
| 630 |
+
"learning_rate": 2.045721369683134e-06,
|
| 631 |
+
"loss": 0.0,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7839721254355401,
|
| 636 |
+
"grad_norm": NaN,
|
| 637 |
+
"learning_rate": 2.045721369683134e-06,
|
| 638 |
+
"loss": 0.0,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7926829268292683,
|
| 643 |
+
"grad_norm": NaN,
|
| 644 |
+
"learning_rate": 2.045721369683134e-06,
|
| 645 |
+
"loss": 0.0,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8013937282229965,
|
| 650 |
+
"grad_norm": NaN,
|
| 651 |
+
"learning_rate": 2.045721369683134e-06,
|
| 652 |
+
"loss": 0.0,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.8101045296167247,
|
| 657 |
+
"grad_norm": NaN,
|
| 658 |
+
"learning_rate": 2.045721369683134e-06,
|
| 659 |
+
"loss": 0.0,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.818815331010453,
|
| 664 |
+
"grad_norm": NaN,
|
| 665 |
+
"learning_rate": 2.045721369683134e-06,
|
| 666 |
+
"loss": 0.0,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.8275261324041812,
|
| 671 |
+
"grad_norm": NaN,
|
| 672 |
+
"learning_rate": 2.045721369683134e-06,
|
| 673 |
+
"loss": 0.0,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.8362369337979094,
|
| 678 |
+
"grad_norm": NaN,
|
| 679 |
+
"learning_rate": 2.045721369683134e-06,
|
| 680 |
+
"loss": 0.0,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.8449477351916377,
|
| 685 |
+
"grad_norm": NaN,
|
| 686 |
+
"learning_rate": 2.045721369683134e-06,
|
| 687 |
+
"loss": 0.0,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8536585365853658,
|
| 692 |
+
"grad_norm": NaN,
|
| 693 |
+
"learning_rate": 2.045721369683134e-06,
|
| 694 |
+
"loss": 0.0,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8623693379790941,
|
| 699 |
+
"grad_norm": NaN,
|
| 700 |
+
"learning_rate": 2.045721369683134e-06,
|
| 701 |
+
"loss": 0.0,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8710801393728222,
|
| 706 |
+
"grad_norm": NaN,
|
| 707 |
+
"learning_rate": 2.045721369683134e-06,
|
| 708 |
+
"loss": 0.0,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8797909407665505,
|
| 713 |
+
"grad_norm": NaN,
|
| 714 |
+
"learning_rate": 2.045721369683134e-06,
|
| 715 |
+
"loss": 0.0,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8885017421602788,
|
| 720 |
+
"grad_norm": NaN,
|
| 721 |
+
"learning_rate": 2.045721369683134e-06,
|
| 722 |
+
"loss": 0.0,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8972125435540069,
|
| 727 |
+
"grad_norm": NaN,
|
| 728 |
+
"learning_rate": 2.045721369683134e-06,
|
| 729 |
+
"loss": 0.0,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9059233449477352,
|
| 734 |
+
"grad_norm": NaN,
|
| 735 |
+
"learning_rate": 2.045721369683134e-06,
|
| 736 |
+
"loss": 0.0,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.9146341463414634,
|
| 741 |
+
"grad_norm": NaN,
|
| 742 |
+
"learning_rate": 2.045721369683134e-06,
|
| 743 |
+
"loss": 0.0,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.9233449477351916,
|
| 748 |
+
"grad_norm": NaN,
|
| 749 |
+
"learning_rate": 2.045721369683134e-06,
|
| 750 |
+
"loss": 0.0,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.9320557491289199,
|
| 755 |
+
"grad_norm": NaN,
|
| 756 |
+
"learning_rate": 2.045721369683134e-06,
|
| 757 |
+
"loss": 0.0,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9407665505226481,
|
| 762 |
+
"grad_norm": NaN,
|
| 763 |
+
"learning_rate": 2.045721369683134e-06,
|
| 764 |
+
"loss": 0.0,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9494773519163763,
|
| 769 |
+
"grad_norm": NaN,
|
| 770 |
+
"learning_rate": 2.045721369683134e-06,
|
| 771 |
+
"loss": 0.0,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9581881533101045,
|
| 776 |
+
"grad_norm": NaN,
|
| 777 |
+
"learning_rate": 2.045721369683134e-06,
|
| 778 |
+
"loss": 0.0,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9668989547038328,
|
| 783 |
+
"grad_norm": NaN,
|
| 784 |
+
"learning_rate": 2.045721369683134e-06,
|
| 785 |
+
"loss": 0.0,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.975609756097561,
|
| 790 |
+
"grad_norm": NaN,
|
| 791 |
+
"learning_rate": 2.045721369683134e-06,
|
| 792 |
+
"loss": 0.0,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9843205574912892,
|
| 797 |
+
"grad_norm": NaN,
|
| 798 |
+
"learning_rate": 2.045721369683134e-06,
|
| 799 |
+
"loss": 0.0,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9930313588850174,
|
| 804 |
+
"grad_norm": NaN,
|
| 805 |
+
"learning_rate": 2.045721369683134e-06,
|
| 806 |
+
"loss": 0.0,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.0017421602787457,
|
| 811 |
+
"grad_norm": NaN,
|
| 812 |
+
"learning_rate": 2.045721369683134e-06,
|
| 813 |
+
"loss": 0.0,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.0104529616724738,
|
| 818 |
+
"grad_norm": NaN,
|
| 819 |
+
"learning_rate": 2.045721369683134e-06,
|
| 820 |
+
"loss": 0.0,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.019163763066202,
|
| 825 |
+
"grad_norm": NaN,
|
| 826 |
+
"learning_rate": 2.045721369683134e-06,
|
| 827 |
+
"loss": 0.0,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.0278745644599303,
|
| 832 |
+
"grad_norm": NaN,
|
| 833 |
+
"learning_rate": 2.045721369683134e-06,
|
| 834 |
+
"loss": 0.0,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.0365853658536586,
|
| 839 |
+
"grad_norm": NaN,
|
| 840 |
+
"learning_rate": 2.045721369683134e-06,
|
| 841 |
+
"loss": 0.0,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.0452961672473868,
|
| 846 |
+
"grad_norm": NaN,
|
| 847 |
+
"learning_rate": 2.045721369683134e-06,
|
| 848 |
+
"loss": 0.0,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.054006968641115,
|
| 853 |
+
"grad_norm": NaN,
|
| 854 |
+
"learning_rate": 2.045721369683134e-06,
|
| 855 |
+
"loss": 0.0,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.0627177700348431,
|
| 860 |
+
"grad_norm": NaN,
|
| 861 |
+
"learning_rate": 2.045721369683134e-06,
|
| 862 |
+
"loss": 0.0,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.0714285714285714,
|
| 867 |
+
"grad_norm": NaN,
|
| 868 |
+
"learning_rate": 2.045721369683134e-06,
|
| 869 |
+
"loss": 0.0,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0801393728222997,
|
| 874 |
+
"grad_norm": NaN,
|
| 875 |
+
"learning_rate": 2.045721369683134e-06,
|
| 876 |
+
"loss": 0.0,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.088850174216028,
|
| 881 |
+
"grad_norm": NaN,
|
| 882 |
+
"learning_rate": 2.045721369683134e-06,
|
| 883 |
+
"loss": 0.0,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.0975609756097562,
|
| 888 |
+
"grad_norm": NaN,
|
| 889 |
+
"learning_rate": 2.045721369683134e-06,
|
| 890 |
+
"loss": 0.0,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.1062717770034842,
|
| 895 |
+
"grad_norm": NaN,
|
| 896 |
+
"learning_rate": 2.045721369683134e-06,
|
| 897 |
+
"loss": 0.0,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.1149825783972125,
|
| 902 |
+
"grad_norm": NaN,
|
| 903 |
+
"learning_rate": 2.045721369683134e-06,
|
| 904 |
+
"loss": 0.0,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.1236933797909407,
|
| 909 |
+
"grad_norm": NaN,
|
| 910 |
+
"learning_rate": 2.045721369683134e-06,
|
| 911 |
+
"loss": 0.0,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.132404181184669,
|
| 916 |
+
"grad_norm": NaN,
|
| 917 |
+
"learning_rate": 2.045721369683134e-06,
|
| 918 |
+
"loss": 0.0,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.1411149825783973,
|
| 923 |
+
"grad_norm": NaN,
|
| 924 |
+
"learning_rate": 2.045721369683134e-06,
|
| 925 |
+
"loss": 0.0,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.1498257839721253,
|
| 930 |
+
"grad_norm": NaN,
|
| 931 |
+
"learning_rate": 2.045721369683134e-06,
|
| 932 |
+
"loss": 0.0,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.1585365853658536,
|
| 937 |
+
"grad_norm": NaN,
|
| 938 |
+
"learning_rate": 2.045721369683134e-06,
|
| 939 |
+
"loss": 0.0,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.1672473867595818,
|
| 944 |
+
"grad_norm": NaN,
|
| 945 |
+
"learning_rate": 2.045721369683134e-06,
|
| 946 |
+
"loss": 0.0,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.17595818815331,
|
| 951 |
+
"grad_norm": NaN,
|
| 952 |
+
"learning_rate": 2.045721369683134e-06,
|
| 953 |
+
"loss": 0.0,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.1846689895470384,
|
| 958 |
+
"grad_norm": NaN,
|
| 959 |
+
"learning_rate": 2.045721369683134e-06,
|
| 960 |
+
"loss": 0.0,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.1933797909407666,
|
| 965 |
+
"grad_norm": NaN,
|
| 966 |
+
"learning_rate": 2.045721369683134e-06,
|
| 967 |
+
"loss": 0.0,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.202090592334495,
|
| 972 |
+
"grad_norm": NaN,
|
| 973 |
+
"learning_rate": 2.045721369683134e-06,
|
| 974 |
+
"loss": 0.0,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.210801393728223,
|
| 979 |
+
"grad_norm": NaN,
|
| 980 |
+
"learning_rate": 2.045721369683134e-06,
|
| 981 |
+
"loss": 0.0,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.2195121951219512,
|
| 986 |
+
"grad_norm": NaN,
|
| 987 |
+
"learning_rate": 2.045721369683134e-06,
|
| 988 |
+
"loss": 0.0,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.2282229965156795,
|
| 993 |
+
"grad_norm": NaN,
|
| 994 |
+
"learning_rate": 2.045721369683134e-06,
|
| 995 |
+
"loss": 0.0,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.2369337979094077,
|
| 1000 |
+
"grad_norm": NaN,
|
| 1001 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1002 |
+
"loss": 0.0,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.245644599303136,
|
| 1007 |
+
"grad_norm": NaN,
|
| 1008 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1009 |
+
"loss": 0.0,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.254355400696864,
|
| 1014 |
+
"grad_norm": NaN,
|
| 1015 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1016 |
+
"loss": 0.0,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.2630662020905923,
|
| 1021 |
+
"grad_norm": NaN,
|
| 1022 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1023 |
+
"loss": 0.0,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.2717770034843205,
|
| 1028 |
+
"grad_norm": NaN,
|
| 1029 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1030 |
+
"loss": 0.0,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.2804878048780488,
|
| 1035 |
+
"grad_norm": NaN,
|
| 1036 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1037 |
+
"loss": 0.0,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.289198606271777,
|
| 1042 |
+
"grad_norm": NaN,
|
| 1043 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1044 |
+
"loss": 0.0,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.297909407665505,
|
| 1049 |
+
"grad_norm": NaN,
|
| 1050 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1051 |
+
"loss": 0.0,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.3066202090592334,
|
| 1056 |
+
"grad_norm": NaN,
|
| 1057 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1058 |
+
"loss": 0.0,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.3153310104529616,
|
| 1063 |
+
"grad_norm": NaN,
|
| 1064 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1065 |
+
"loss": 0.0,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.32404181184669,
|
| 1070 |
+
"grad_norm": NaN,
|
| 1071 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1072 |
+
"loss": 0.0,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.3327526132404182,
|
| 1077 |
+
"grad_norm": NaN,
|
| 1078 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1079 |
+
"loss": 0.0,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.3414634146341464,
|
| 1084 |
+
"grad_norm": NaN,
|
| 1085 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1086 |
+
"loss": 0.0,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.3501742160278747,
|
| 1091 |
+
"grad_norm": NaN,
|
| 1092 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1093 |
+
"loss": 0.0,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.3588850174216027,
|
| 1098 |
+
"grad_norm": NaN,
|
| 1099 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1100 |
+
"loss": 0.0,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.367595818815331,
|
| 1105 |
+
"grad_norm": NaN,
|
| 1106 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1107 |
+
"loss": 0.0,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3763066202090593,
|
| 1112 |
+
"grad_norm": NaN,
|
| 1113 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1114 |
+
"loss": 0.0,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.3850174216027875,
|
| 1119 |
+
"grad_norm": NaN,
|
| 1120 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1121 |
+
"loss": 0.0,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.3937282229965158,
|
| 1126 |
+
"grad_norm": NaN,
|
| 1127 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1128 |
+
"loss": 0.0,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.4024390243902438,
|
| 1133 |
+
"grad_norm": NaN,
|
| 1134 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1135 |
+
"loss": 0.0,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.411149825783972,
|
| 1140 |
+
"grad_norm": NaN,
|
| 1141 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1142 |
+
"loss": 0.0,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.4198606271777003,
|
| 1147 |
+
"grad_norm": NaN,
|
| 1148 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1149 |
+
"loss": 0.0,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.4285714285714286,
|
| 1154 |
+
"grad_norm": NaN,
|
| 1155 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1156 |
+
"loss": 0.0,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.4372822299651569,
|
| 1161 |
+
"grad_norm": NaN,
|
| 1162 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1163 |
+
"loss": 0.0,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.445993031358885,
|
| 1168 |
+
"grad_norm": NaN,
|
| 1169 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1170 |
+
"loss": 0.0,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.4547038327526132,
|
| 1175 |
+
"grad_norm": NaN,
|
| 1176 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1177 |
+
"loss": 0.0,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.4634146341463414,
|
| 1182 |
+
"grad_norm": NaN,
|
| 1183 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1184 |
+
"loss": 0.0,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.4721254355400697,
|
| 1189 |
+
"grad_norm": NaN,
|
| 1190 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1191 |
+
"loss": 0.0,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.480836236933798,
|
| 1196 |
+
"grad_norm": NaN,
|
| 1197 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1198 |
+
"loss": 0.0,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.489547038327526,
|
| 1203 |
+
"grad_norm": NaN,
|
| 1204 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1205 |
+
"loss": 0.0,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.4982578397212545,
|
| 1210 |
+
"grad_norm": NaN,
|
| 1211 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1212 |
+
"loss": 0.0,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.5069686411149825,
|
| 1217 |
+
"grad_norm": NaN,
|
| 1218 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1219 |
+
"loss": 0.0,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.5156794425087108,
|
| 1224 |
+
"grad_norm": NaN,
|
| 1225 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1226 |
+
"loss": 0.0,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.524390243902439,
|
| 1231 |
+
"grad_norm": NaN,
|
| 1232 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1233 |
+
"loss": 0.0,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.533101045296167,
|
| 1238 |
+
"grad_norm": NaN,
|
| 1239 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1240 |
+
"loss": 0.0,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.5418118466898956,
|
| 1245 |
+
"grad_norm": NaN,
|
| 1246 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1247 |
+
"loss": 0.0,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.5505226480836236,
|
| 1252 |
+
"grad_norm": NaN,
|
| 1253 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1254 |
+
"loss": 0.0,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.5592334494773519,
|
| 1259 |
+
"grad_norm": NaN,
|
| 1260 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1261 |
+
"loss": 0.0,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.5679442508710801,
|
| 1266 |
+
"grad_norm": NaN,
|
| 1267 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1268 |
+
"loss": 0.0,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.5766550522648084,
|
| 1273 |
+
"grad_norm": NaN,
|
| 1274 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1275 |
+
"loss": 0.0,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.5853658536585367,
|
| 1280 |
+
"grad_norm": NaN,
|
| 1281 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1282 |
+
"loss": 0.0,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.5940766550522647,
|
| 1287 |
+
"grad_norm": NaN,
|
| 1288 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1289 |
+
"loss": 0.0,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.6027874564459932,
|
| 1294 |
+
"grad_norm": NaN,
|
| 1295 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1296 |
+
"loss": 0.0,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.6114982578397212,
|
| 1301 |
+
"grad_norm": NaN,
|
| 1302 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1303 |
+
"loss": 0.0,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.6202090592334495,
|
| 1308 |
+
"grad_norm": NaN,
|
| 1309 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1310 |
+
"loss": 0.0,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.6289198606271778,
|
| 1315 |
+
"grad_norm": NaN,
|
| 1316 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1317 |
+
"loss": 0.0,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.6376306620209058,
|
| 1322 |
+
"grad_norm": NaN,
|
| 1323 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1324 |
+
"loss": 0.0,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.6463414634146343,
|
| 1329 |
+
"grad_norm": NaN,
|
| 1330 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1331 |
+
"loss": 0.0,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.6550522648083623,
|
| 1336 |
+
"grad_norm": NaN,
|
| 1337 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1338 |
+
"loss": 0.0,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.6637630662020906,
|
| 1343 |
+
"grad_norm": NaN,
|
| 1344 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1345 |
+
"loss": 0.0,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.6724738675958188,
|
| 1350 |
+
"grad_norm": NaN,
|
| 1351 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1352 |
+
"loss": 0.0,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.6811846689895469,
|
| 1357 |
+
"grad_norm": NaN,
|
| 1358 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1359 |
+
"loss": 0.0,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.6898954703832754,
|
| 1364 |
+
"grad_norm": NaN,
|
| 1365 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1366 |
+
"loss": 0.0,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.6986062717770034,
|
| 1371 |
+
"grad_norm": NaN,
|
| 1372 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1373 |
+
"loss": 0.0,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.7073170731707317,
|
| 1378 |
+
"grad_norm": NaN,
|
| 1379 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1380 |
+
"loss": 0.0,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.71602787456446,
|
| 1385 |
+
"grad_norm": NaN,
|
| 1386 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1387 |
+
"loss": 0.0,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.7247386759581882,
|
| 1392 |
+
"grad_norm": NaN,
|
| 1393 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1394 |
+
"loss": 0.0,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.7334494773519165,
|
| 1399 |
+
"grad_norm": NaN,
|
| 1400 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1401 |
+
"loss": 0.0,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.7421602787456445,
|
| 1406 |
+
"grad_norm": NaN,
|
| 1407 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1408 |
+
"loss": 0.0,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.750871080139373,
|
| 1413 |
+
"grad_norm": NaN,
|
| 1414 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1415 |
+
"loss": 0.0,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.759581881533101,
|
| 1420 |
+
"grad_norm": NaN,
|
| 1421 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1422 |
+
"loss": 0.0,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.7682926829268293,
|
| 1427 |
+
"grad_norm": NaN,
|
| 1428 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1429 |
+
"loss": 0.0,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.7770034843205575,
|
| 1434 |
+
"grad_norm": NaN,
|
| 1435 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1436 |
+
"loss": 0.0,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.7857142857142856,
|
| 1441 |
+
"grad_norm": NaN,
|
| 1442 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1443 |
+
"loss": 0.0,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.794425087108014,
|
| 1448 |
+
"grad_norm": NaN,
|
| 1449 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1450 |
+
"loss": 0.0,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.8031358885017421,
|
| 1455 |
+
"grad_norm": NaN,
|
| 1456 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1457 |
+
"loss": 0.0,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.8118466898954704,
|
| 1462 |
+
"grad_norm": NaN,
|
| 1463 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1464 |
+
"loss": 0.0,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.8205574912891986,
|
| 1469 |
+
"grad_norm": NaN,
|
| 1470 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1471 |
+
"loss": 0.0,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.8292682926829267,
|
| 1476 |
+
"grad_norm": NaN,
|
| 1477 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1478 |
+
"loss": 0.0,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.8379790940766552,
|
| 1483 |
+
"grad_norm": NaN,
|
| 1484 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1485 |
+
"loss": 0.0,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.8466898954703832,
|
| 1490 |
+
"grad_norm": NaN,
|
| 1491 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1492 |
+
"loss": 0.0,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.8554006968641115,
|
| 1497 |
+
"grad_norm": NaN,
|
| 1498 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1499 |
+
"loss": 0.0,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.8641114982578397,
|
| 1504 |
+
"grad_norm": NaN,
|
| 1505 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1506 |
+
"loss": 0.0,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.872822299651568,
|
| 1511 |
+
"grad_norm": NaN,
|
| 1512 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1513 |
+
"loss": 0.0,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.8815331010452963,
|
| 1518 |
+
"grad_norm": NaN,
|
| 1519 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1520 |
+
"loss": 0.0,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.8902439024390243,
|
| 1525 |
+
"grad_norm": NaN,
|
| 1526 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1527 |
+
"loss": 0.0,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.8989547038327528,
|
| 1532 |
+
"grad_norm": NaN,
|
| 1533 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1534 |
+
"loss": 0.0,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.9076655052264808,
|
| 1539 |
+
"grad_norm": NaN,
|
| 1540 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1541 |
+
"loss": 0.0,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.916376306620209,
|
| 1546 |
+
"grad_norm": NaN,
|
| 1547 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1548 |
+
"loss": 0.0,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.9250871080139373,
|
| 1553 |
+
"grad_norm": NaN,
|
| 1554 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1555 |
+
"loss": 0.0,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.9337979094076654,
|
| 1560 |
+
"grad_norm": NaN,
|
| 1561 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1562 |
+
"loss": 0.0,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.9425087108013939,
|
| 1567 |
+
"grad_norm": NaN,
|
| 1568 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1569 |
+
"loss": 0.0,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.951219512195122,
|
| 1574 |
+
"grad_norm": NaN,
|
| 1575 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1576 |
+
"loss": 0.0,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.9599303135888502,
|
| 1581 |
+
"grad_norm": NaN,
|
| 1582 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1583 |
+
"loss": 0.0,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.9686411149825784,
|
| 1588 |
+
"grad_norm": NaN,
|
| 1589 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1590 |
+
"loss": 0.0,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.9773519163763065,
|
| 1595 |
+
"grad_norm": NaN,
|
| 1596 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1597 |
+
"loss": 0.0,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.986062717770035,
|
| 1602 |
+
"grad_norm": NaN,
|
| 1603 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1604 |
+
"loss": 0.0,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.994773519163763,
|
| 1609 |
+
"grad_norm": NaN,
|
| 1610 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1611 |
+
"loss": 0.0,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.0034843205574915,
|
| 1616 |
+
"grad_norm": NaN,
|
| 1617 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1618 |
+
"loss": 0.0,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.0121951219512195,
|
| 1623 |
+
"grad_norm": NaN,
|
| 1624 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1625 |
+
"loss": 0.0,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.0209059233449476,
|
| 1630 |
+
"grad_norm": NaN,
|
| 1631 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1632 |
+
"loss": 0.0,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.029616724738676,
|
| 1637 |
+
"grad_norm": NaN,
|
| 1638 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1639 |
+
"loss": 0.0,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.038327526132404,
|
| 1644 |
+
"grad_norm": NaN,
|
| 1645 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1646 |
+
"loss": 0.0,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.0470383275261326,
|
| 1651 |
+
"grad_norm": NaN,
|
| 1652 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1653 |
+
"loss": 0.0,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.0557491289198606,
|
| 1658 |
+
"grad_norm": NaN,
|
| 1659 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1660 |
+
"loss": 0.0,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.0644599303135887,
|
| 1665 |
+
"grad_norm": NaN,
|
| 1666 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1667 |
+
"loss": 0.0,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.073170731707317,
|
| 1672 |
+
"grad_norm": NaN,
|
| 1673 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1674 |
+
"loss": 0.0,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.081881533101045,
|
| 1679 |
+
"grad_norm": NaN,
|
| 1680 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1681 |
+
"loss": 0.0,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.0905923344947737,
|
| 1686 |
+
"grad_norm": NaN,
|
| 1687 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1688 |
+
"loss": 0.0,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.0993031358885017,
|
| 1693 |
+
"grad_norm": NaN,
|
| 1694 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1695 |
+
"loss": 0.0,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.10801393728223,
|
| 1700 |
+
"grad_norm": NaN,
|
| 1701 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1702 |
+
"loss": 0.0,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.1167247386759582,
|
| 1707 |
+
"grad_norm": NaN,
|
| 1708 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1709 |
+
"loss": 0.0,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.1254355400696863,
|
| 1714 |
+
"grad_norm": NaN,
|
| 1715 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1716 |
+
"loss": 0.0,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.1341463414634148,
|
| 1721 |
+
"grad_norm": NaN,
|
| 1722 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1723 |
+
"loss": 0.0,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.142857142857143,
|
| 1728 |
+
"grad_norm": NaN,
|
| 1729 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1730 |
+
"loss": 0.0,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.1515679442508713,
|
| 1735 |
+
"grad_norm": NaN,
|
| 1736 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1737 |
+
"loss": 0.0,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.1602787456445993,
|
| 1742 |
+
"grad_norm": NaN,
|
| 1743 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1744 |
+
"loss": 0.0,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.1689895470383274,
|
| 1749 |
+
"grad_norm": NaN,
|
| 1750 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1751 |
+
"loss": 0.0,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.177700348432056,
|
| 1756 |
+
"grad_norm": NaN,
|
| 1757 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1758 |
+
"loss": 0.0,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.186411149825784,
|
| 1763 |
+
"grad_norm": NaN,
|
| 1764 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1765 |
+
"loss": 0.0,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.1951219512195124,
|
| 1770 |
+
"grad_norm": NaN,
|
| 1771 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1772 |
+
"loss": 0.0,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.2038327526132404,
|
| 1777 |
+
"grad_norm": NaN,
|
| 1778 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1779 |
+
"loss": 0.0,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.2125435540069684,
|
| 1784 |
+
"grad_norm": NaN,
|
| 1785 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1786 |
+
"loss": 0.0,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.221254355400697,
|
| 1791 |
+
"grad_norm": NaN,
|
| 1792 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1793 |
+
"loss": 0.0,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.229965156794425,
|
| 1798 |
+
"grad_norm": NaN,
|
| 1799 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1800 |
+
"loss": 0.0,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.2386759581881535,
|
| 1805 |
+
"grad_norm": NaN,
|
| 1806 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1807 |
+
"loss": 0.0,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.2473867595818815,
|
| 1812 |
+
"grad_norm": NaN,
|
| 1813 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1814 |
+
"loss": 0.0,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.2560975609756095,
|
| 1819 |
+
"grad_norm": NaN,
|
| 1820 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1821 |
+
"loss": 0.0,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.264808362369338,
|
| 1826 |
+
"grad_norm": NaN,
|
| 1827 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1828 |
+
"loss": 0.0,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.273519163763066,
|
| 1833 |
+
"grad_norm": NaN,
|
| 1834 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1835 |
+
"loss": 0.0,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.2822299651567945,
|
| 1840 |
+
"grad_norm": NaN,
|
| 1841 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1842 |
+
"loss": 0.0,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.2909407665505226,
|
| 1847 |
+
"grad_norm": NaN,
|
| 1848 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1849 |
+
"loss": 0.0,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.2996515679442506,
|
| 1854 |
+
"grad_norm": NaN,
|
| 1855 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1856 |
+
"loss": 0.0,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.308362369337979,
|
| 1861 |
+
"grad_norm": NaN,
|
| 1862 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1863 |
+
"loss": 0.0,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.317073170731707,
|
| 1868 |
+
"grad_norm": NaN,
|
| 1869 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1870 |
+
"loss": 0.0,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.3257839721254356,
|
| 1875 |
+
"grad_norm": NaN,
|
| 1876 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1877 |
+
"loss": 0.0,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.3344947735191637,
|
| 1882 |
+
"grad_norm": NaN,
|
| 1883 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1884 |
+
"loss": 0.0,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.343205574912892,
|
| 1889 |
+
"grad_norm": NaN,
|
| 1890 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1891 |
+
"loss": 0.0,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.35191637630662,
|
| 1896 |
+
"grad_norm": NaN,
|
| 1897 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1898 |
+
"loss": 0.0,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.3606271777003482,
|
| 1903 |
+
"grad_norm": NaN,
|
| 1904 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1905 |
+
"loss": 0.0,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.3693379790940767,
|
| 1910 |
+
"grad_norm": NaN,
|
| 1911 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1912 |
+
"loss": 0.0,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.3780487804878048,
|
| 1917 |
+
"grad_norm": NaN,
|
| 1918 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1919 |
+
"loss": 0.0,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.3867595818815333,
|
| 1924 |
+
"grad_norm": NaN,
|
| 1925 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1926 |
+
"loss": 0.0,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.3954703832752613,
|
| 1931 |
+
"grad_norm": NaN,
|
| 1932 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1933 |
+
"loss": 0.0,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.40418118466899,
|
| 1938 |
+
"grad_norm": NaN,
|
| 1939 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1940 |
+
"loss": 0.0,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.412891986062718,
|
| 1945 |
+
"grad_norm": NaN,
|
| 1946 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1947 |
+
"loss": 0.0,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.421602787456446,
|
| 1952 |
+
"grad_norm": NaN,
|
| 1953 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1954 |
+
"loss": 0.0,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.4303135888501743,
|
| 1959 |
+
"grad_norm": NaN,
|
| 1960 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1961 |
+
"loss": 0.0,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.4390243902439024,
|
| 1966 |
+
"grad_norm": NaN,
|
| 1967 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1968 |
+
"loss": 0.0,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.447735191637631,
|
| 1973 |
+
"grad_norm": NaN,
|
| 1974 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1975 |
+
"loss": 0.0,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.456445993031359,
|
| 1980 |
+
"grad_norm": NaN,
|
| 1981 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1982 |
+
"loss": 0.0,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.465156794425087,
|
| 1987 |
+
"grad_norm": NaN,
|
| 1988 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1989 |
+
"loss": 0.0,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.4738675958188154,
|
| 1994 |
+
"grad_norm": NaN,
|
| 1995 |
+
"learning_rate": 2.045721369683134e-06,
|
| 1996 |
+
"loss": 0.0,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.4825783972125435,
|
| 2001 |
+
"grad_norm": NaN,
|
| 2002 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2003 |
+
"loss": 0.0,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.491289198606272,
|
| 2008 |
+
"grad_norm": NaN,
|
| 2009 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2010 |
+
"loss": 0.0,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.5,
|
| 2015 |
+
"grad_norm": NaN,
|
| 2016 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2017 |
+
"loss": 0.0,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.508710801393728,
|
| 2022 |
+
"grad_norm": NaN,
|
| 2023 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2024 |
+
"loss": 0.0,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.5174216027874565,
|
| 2029 |
+
"grad_norm": NaN,
|
| 2030 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2031 |
+
"loss": 0.0,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.5261324041811846,
|
| 2036 |
+
"grad_norm": NaN,
|
| 2037 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2038 |
+
"loss": 0.0,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.534843205574913,
|
| 2043 |
+
"grad_norm": NaN,
|
| 2044 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2045 |
+
"loss": 0.0,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.543554006968641,
|
| 2050 |
+
"grad_norm": NaN,
|
| 2051 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2052 |
+
"loss": 0.0,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.552264808362369,
|
| 2057 |
+
"grad_norm": NaN,
|
| 2058 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2059 |
+
"loss": 0.0,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.5609756097560976,
|
| 2064 |
+
"grad_norm": NaN,
|
| 2065 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2066 |
+
"loss": 0.0,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.5696864111498257,
|
| 2071 |
+
"grad_norm": NaN,
|
| 2072 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2073 |
+
"loss": 0.0,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.578397212543554,
|
| 2078 |
+
"grad_norm": NaN,
|
| 2079 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2080 |
+
"loss": 0.0,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.587108013937282,
|
| 2085 |
+
"grad_norm": NaN,
|
| 2086 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2087 |
+
"loss": 0.0,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.59581881533101,
|
| 2092 |
+
"grad_norm": NaN,
|
| 2093 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2094 |
+
"loss": 0.0,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.6045296167247387,
|
| 2099 |
+
"grad_norm": NaN,
|
| 2100 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2101 |
+
"loss": 0.0,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.6132404181184667,
|
| 2106 |
+
"grad_norm": NaN,
|
| 2107 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2108 |
+
"loss": 0.0,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.6219512195121952,
|
| 2113 |
+
"grad_norm": NaN,
|
| 2114 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2115 |
+
"loss": 0.0,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.6306620209059233,
|
| 2120 |
+
"grad_norm": NaN,
|
| 2121 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2122 |
+
"loss": 0.0,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.6393728222996513,
|
| 2127 |
+
"grad_norm": NaN,
|
| 2128 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2129 |
+
"loss": 0.0,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.64808362369338,
|
| 2134 |
+
"grad_norm": NaN,
|
| 2135 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2136 |
+
"loss": 0.0,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.6567944250871083,
|
| 2141 |
+
"grad_norm": NaN,
|
| 2142 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2143 |
+
"loss": 0.0,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.6655052264808363,
|
| 2148 |
+
"grad_norm": NaN,
|
| 2149 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2150 |
+
"loss": 0.0,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.6742160278745644,
|
| 2155 |
+
"grad_norm": NaN,
|
| 2156 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2157 |
+
"loss": 0.0,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.682926829268293,
|
| 2162 |
+
"grad_norm": NaN,
|
| 2163 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2164 |
+
"loss": 0.0,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.691637630662021,
|
| 2169 |
+
"grad_norm": NaN,
|
| 2170 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2171 |
+
"loss": 0.0,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.7003484320557494,
|
| 2176 |
+
"grad_norm": NaN,
|
| 2177 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2178 |
+
"loss": 0.0,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.7090592334494774,
|
| 2183 |
+
"grad_norm": NaN,
|
| 2184 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2185 |
+
"loss": 0.0,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.7177700348432055,
|
| 2190 |
+
"grad_norm": NaN,
|
| 2191 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2192 |
+
"loss": 0.0,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.726480836236934,
|
| 2197 |
+
"grad_norm": NaN,
|
| 2198 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2199 |
+
"loss": 0.0,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.735191637630662,
|
| 2204 |
+
"grad_norm": NaN,
|
| 2205 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2206 |
+
"loss": 0.0,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.7439024390243905,
|
| 2211 |
+
"grad_norm": NaN,
|
| 2212 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2213 |
+
"loss": 0.0,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.7526132404181185,
|
| 2218 |
+
"grad_norm": NaN,
|
| 2219 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2220 |
+
"loss": 0.0,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.7613240418118465,
|
| 2225 |
+
"grad_norm": NaN,
|
| 2226 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2227 |
+
"loss": 0.0,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.770034843205575,
|
| 2232 |
+
"grad_norm": NaN,
|
| 2233 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2234 |
+
"loss": 0.0,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.778745644599303,
|
| 2239 |
+
"grad_norm": NaN,
|
| 2240 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2241 |
+
"loss": 0.0,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.7874564459930316,
|
| 2246 |
+
"grad_norm": NaN,
|
| 2247 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2248 |
+
"loss": 0.0,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.7961672473867596,
|
| 2253 |
+
"grad_norm": NaN,
|
| 2254 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2255 |
+
"loss": 0.0,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.8048780487804876,
|
| 2260 |
+
"grad_norm": NaN,
|
| 2261 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2262 |
+
"loss": 0.0,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.813588850174216,
|
| 2267 |
+
"grad_norm": NaN,
|
| 2268 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2269 |
+
"loss": 0.0,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.822299651567944,
|
| 2274 |
+
"grad_norm": NaN,
|
| 2275 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2276 |
+
"loss": 0.0,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.8310104529616726,
|
| 2281 |
+
"grad_norm": NaN,
|
| 2282 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2283 |
+
"loss": 0.0,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.8397212543554007,
|
| 2288 |
+
"grad_norm": NaN,
|
| 2289 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2290 |
+
"loss": 0.0,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.8484320557491287,
|
| 2295 |
+
"grad_norm": NaN,
|
| 2296 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2297 |
+
"loss": 0.0,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.857142857142857,
|
| 2302 |
+
"grad_norm": NaN,
|
| 2303 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2304 |
+
"loss": 0.0,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.8658536585365852,
|
| 2309 |
+
"grad_norm": NaN,
|
| 2310 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2311 |
+
"loss": 0.0,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.8745644599303137,
|
| 2316 |
+
"grad_norm": NaN,
|
| 2317 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2318 |
+
"loss": 0.0,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.8832752613240418,
|
| 2323 |
+
"grad_norm": NaN,
|
| 2324 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2325 |
+
"loss": 0.0,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.89198606271777,
|
| 2330 |
+
"grad_norm": NaN,
|
| 2331 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2332 |
+
"loss": 0.0,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.9006968641114983,
|
| 2337 |
+
"grad_norm": NaN,
|
| 2338 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2339 |
+
"loss": 0.0,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.9094076655052263,
|
| 2344 |
+
"grad_norm": NaN,
|
| 2345 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2346 |
+
"loss": 0.0,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.918118466898955,
|
| 2351 |
+
"grad_norm": NaN,
|
| 2352 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2353 |
+
"loss": 0.0,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.926829268292683,
|
| 2358 |
+
"grad_norm": NaN,
|
| 2359 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2360 |
+
"loss": 0.0,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.935540069686411,
|
| 2365 |
+
"grad_norm": NaN,
|
| 2366 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2367 |
+
"loss": 0.0,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.9442508710801394,
|
| 2372 |
+
"grad_norm": NaN,
|
| 2373 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2374 |
+
"loss": 0.0,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.952961672473868,
|
| 2379 |
+
"grad_norm": NaN,
|
| 2380 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2381 |
+
"loss": 0.0,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.961672473867596,
|
| 2386 |
+
"grad_norm": NaN,
|
| 2387 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2388 |
+
"loss": 0.0,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.970383275261324,
|
| 2393 |
+
"grad_norm": NaN,
|
| 2394 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2395 |
+
"loss": 0.0,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.979094076655052,
|
| 2400 |
+
"grad_norm": NaN,
|
| 2401 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2402 |
+
"loss": 0.0,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.9878048780487805,
|
| 2407 |
+
"grad_norm": NaN,
|
| 2408 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2409 |
+
"loss": 0.0,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.996515679442509,
|
| 2414 |
+
"grad_norm": NaN,
|
| 2415 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2416 |
+
"loss": 0.0,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 3.005226480836237,
|
| 2421 |
+
"grad_norm": NaN,
|
| 2422 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2423 |
+
"loss": 0.0,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 3.013937282229965,
|
| 2428 |
+
"grad_norm": NaN,
|
| 2429 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2430 |
+
"loss": 0.0,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 3.0226480836236935,
|
| 2435 |
+
"grad_norm": NaN,
|
| 2436 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2437 |
+
"loss": 0.0,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 3.0313588850174216,
|
| 2442 |
+
"grad_norm": NaN,
|
| 2443 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2444 |
+
"loss": 0.0,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 3.0400696864111496,
|
| 2449 |
+
"grad_norm": NaN,
|
| 2450 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2451 |
+
"loss": 0.0,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 3.048780487804878,
|
| 2456 |
+
"grad_norm": NaN,
|
| 2457 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2458 |
+
"loss": 0.0,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 3.057491289198606,
|
| 2463 |
+
"grad_norm": NaN,
|
| 2464 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2465 |
+
"loss": 0.0,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 3.0662020905923346,
|
| 2470 |
+
"grad_norm": NaN,
|
| 2471 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2472 |
+
"loss": 0.0,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 3.0749128919860627,
|
| 2477 |
+
"grad_norm": NaN,
|
| 2478 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2479 |
+
"loss": 0.0,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 3.083623693379791,
|
| 2484 |
+
"grad_norm": NaN,
|
| 2485 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2486 |
+
"loss": 0.0,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 3.092334494773519,
|
| 2491 |
+
"grad_norm": NaN,
|
| 2492 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2493 |
+
"loss": 0.0,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 3.1010452961672472,
|
| 2498 |
+
"grad_norm": NaN,
|
| 2499 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2500 |
+
"loss": 0.0,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 3.1097560975609757,
|
| 2505 |
+
"grad_norm": NaN,
|
| 2506 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2507 |
+
"loss": 0.0,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 3.1184668989547037,
|
| 2512 |
+
"grad_norm": NaN,
|
| 2513 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2514 |
+
"loss": 0.0,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 3.1271777003484322,
|
| 2519 |
+
"grad_norm": NaN,
|
| 2520 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2521 |
+
"loss": 0.0,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 3.1358885017421603,
|
| 2526 |
+
"grad_norm": NaN,
|
| 2527 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2528 |
+
"loss": 0.0,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 3.1445993031358883,
|
| 2533 |
+
"grad_norm": NaN,
|
| 2534 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2535 |
+
"loss": 0.0,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 3.153310104529617,
|
| 2540 |
+
"grad_norm": NaN,
|
| 2541 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2542 |
+
"loss": 0.0,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 3.162020905923345,
|
| 2547 |
+
"grad_norm": NaN,
|
| 2548 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2549 |
+
"loss": 0.0,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 3.1707317073170733,
|
| 2554 |
+
"grad_norm": NaN,
|
| 2555 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2556 |
+
"loss": 0.0,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 3.1794425087108014,
|
| 2561 |
+
"grad_norm": NaN,
|
| 2562 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2563 |
+
"loss": 0.0,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 3.1881533101045294,
|
| 2568 |
+
"grad_norm": NaN,
|
| 2569 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2570 |
+
"loss": 0.0,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 3.196864111498258,
|
| 2575 |
+
"grad_norm": NaN,
|
| 2576 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2577 |
+
"loss": 0.0,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 3.205574912891986,
|
| 2582 |
+
"grad_norm": NaN,
|
| 2583 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2584 |
+
"loss": 0.0,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 3.2142857142857144,
|
| 2589 |
+
"grad_norm": NaN,
|
| 2590 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2591 |
+
"loss": 0.0,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 3.2229965156794425,
|
| 2596 |
+
"grad_norm": NaN,
|
| 2597 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2598 |
+
"loss": 0.0,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 3.231707317073171,
|
| 2603 |
+
"grad_norm": NaN,
|
| 2604 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2605 |
+
"loss": 0.0,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 3.240418118466899,
|
| 2610 |
+
"grad_norm": NaN,
|
| 2611 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2612 |
+
"loss": 0.0,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 3.249128919860627,
|
| 2617 |
+
"grad_norm": NaN,
|
| 2618 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2619 |
+
"loss": 0.0,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 3.2578397212543555,
|
| 2624 |
+
"grad_norm": NaN,
|
| 2625 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2626 |
+
"loss": 0.0,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 3.2665505226480835,
|
| 2631 |
+
"grad_norm": NaN,
|
| 2632 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2633 |
+
"loss": 0.0,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 3.275261324041812,
|
| 2638 |
+
"grad_norm": NaN,
|
| 2639 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2640 |
+
"loss": 0.0,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 3.28397212543554,
|
| 2645 |
+
"grad_norm": NaN,
|
| 2646 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2647 |
+
"loss": 0.0,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 3.292682926829268,
|
| 2652 |
+
"grad_norm": NaN,
|
| 2653 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2654 |
+
"loss": 0.0,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 3.3013937282229966,
|
| 2659 |
+
"grad_norm": NaN,
|
| 2660 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2661 |
+
"loss": 0.0,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 3.3101045296167246,
|
| 2666 |
+
"grad_norm": NaN,
|
| 2667 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2668 |
+
"loss": 0.0,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 3.318815331010453,
|
| 2673 |
+
"grad_norm": NaN,
|
| 2674 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2675 |
+
"loss": 0.0,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 3.327526132404181,
|
| 2680 |
+
"grad_norm": NaN,
|
| 2681 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2682 |
+
"loss": 0.0,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 3.3362369337979096,
|
| 2687 |
+
"grad_norm": NaN,
|
| 2688 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2689 |
+
"loss": 0.0,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 3.3449477351916377,
|
| 2694 |
+
"grad_norm": NaN,
|
| 2695 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2696 |
+
"loss": 0.0,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 3.3536585365853657,
|
| 2701 |
+
"grad_norm": NaN,
|
| 2702 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2703 |
+
"loss": 0.0,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 3.362369337979094,
|
| 2708 |
+
"grad_norm": NaN,
|
| 2709 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2710 |
+
"loss": 0.0,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 3.3710801393728222,
|
| 2715 |
+
"grad_norm": NaN,
|
| 2716 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2717 |
+
"loss": 0.0,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 3.3797909407665507,
|
| 2722 |
+
"grad_norm": NaN,
|
| 2723 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2724 |
+
"loss": 0.0,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 3.3885017421602788,
|
| 2729 |
+
"grad_norm": NaN,
|
| 2730 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2731 |
+
"loss": 0.0,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 3.397212543554007,
|
| 2736 |
+
"grad_norm": NaN,
|
| 2737 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2738 |
+
"loss": 0.0,
|
| 2739 |
+
"step": 3900
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 3.4059233449477353,
|
| 2743 |
+
"grad_norm": NaN,
|
| 2744 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2745 |
+
"loss": 0.0,
|
| 2746 |
+
"step": 3910
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 3.4146341463414633,
|
| 2750 |
+
"grad_norm": NaN,
|
| 2751 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2752 |
+
"loss": 0.0,
|
| 2753 |
+
"step": 3920
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 3.423344947735192,
|
| 2757 |
+
"grad_norm": NaN,
|
| 2758 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2759 |
+
"loss": 0.0,
|
| 2760 |
+
"step": 3930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 3.43205574912892,
|
| 2764 |
+
"grad_norm": NaN,
|
| 2765 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2766 |
+
"loss": 0.0,
|
| 2767 |
+
"step": 3940
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 3.440766550522648,
|
| 2771 |
+
"grad_norm": NaN,
|
| 2772 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2773 |
+
"loss": 0.0,
|
| 2774 |
+
"step": 3950
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 3.4494773519163764,
|
| 2778 |
+
"grad_norm": NaN,
|
| 2779 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2780 |
+
"loss": 0.0,
|
| 2781 |
+
"step": 3960
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 3.4581881533101044,
|
| 2785 |
+
"grad_norm": NaN,
|
| 2786 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2787 |
+
"loss": 0.0,
|
| 2788 |
+
"step": 3970
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 3.466898954703833,
|
| 2792 |
+
"grad_norm": NaN,
|
| 2793 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2794 |
+
"loss": 0.0,
|
| 2795 |
+
"step": 3980
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 3.475609756097561,
|
| 2799 |
+
"grad_norm": NaN,
|
| 2800 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2801 |
+
"loss": 0.0,
|
| 2802 |
+
"step": 3990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 3.484320557491289,
|
| 2806 |
+
"grad_norm": NaN,
|
| 2807 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2808 |
+
"loss": 0.0,
|
| 2809 |
+
"step": 4000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 3.4930313588850175,
|
| 2813 |
+
"grad_norm": NaN,
|
| 2814 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2815 |
+
"loss": 0.0,
|
| 2816 |
+
"step": 4010
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 3.5017421602787455,
|
| 2820 |
+
"grad_norm": NaN,
|
| 2821 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2822 |
+
"loss": 0.0,
|
| 2823 |
+
"step": 4020
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 3.510452961672474,
|
| 2827 |
+
"grad_norm": NaN,
|
| 2828 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2829 |
+
"loss": 0.0,
|
| 2830 |
+
"step": 4030
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 3.519163763066202,
|
| 2834 |
+
"grad_norm": NaN,
|
| 2835 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2836 |
+
"loss": 0.0,
|
| 2837 |
+
"step": 4040
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 3.52787456445993,
|
| 2841 |
+
"grad_norm": NaN,
|
| 2842 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2843 |
+
"loss": 0.0,
|
| 2844 |
+
"step": 4050
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 3.5365853658536586,
|
| 2848 |
+
"grad_norm": NaN,
|
| 2849 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2850 |
+
"loss": 0.0,
|
| 2851 |
+
"step": 4060
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 3.5452961672473866,
|
| 2855 |
+
"grad_norm": NaN,
|
| 2856 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2857 |
+
"loss": 0.0,
|
| 2858 |
+
"step": 4070
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 3.554006968641115,
|
| 2862 |
+
"grad_norm": NaN,
|
| 2863 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2864 |
+
"loss": 0.0,
|
| 2865 |
+
"step": 4080
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 3.562717770034843,
|
| 2869 |
+
"grad_norm": NaN,
|
| 2870 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2871 |
+
"loss": 0.0,
|
| 2872 |
+
"step": 4090
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 3.571428571428571,
|
| 2876 |
+
"grad_norm": NaN,
|
| 2877 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2878 |
+
"loss": 0.0,
|
| 2879 |
+
"step": 4100
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 3.5801393728222997,
|
| 2883 |
+
"grad_norm": NaN,
|
| 2884 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2885 |
+
"loss": 0.0,
|
| 2886 |
+
"step": 4110
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 3.588850174216028,
|
| 2890 |
+
"grad_norm": NaN,
|
| 2891 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2892 |
+
"loss": 0.0,
|
| 2893 |
+
"step": 4120
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 3.597560975609756,
|
| 2897 |
+
"grad_norm": NaN,
|
| 2898 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2899 |
+
"loss": 0.0,
|
| 2900 |
+
"step": 4130
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 3.6062717770034842,
|
| 2904 |
+
"grad_norm": NaN,
|
| 2905 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2906 |
+
"loss": 0.0,
|
| 2907 |
+
"step": 4140
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 3.6149825783972127,
|
| 2911 |
+
"grad_norm": NaN,
|
| 2912 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2913 |
+
"loss": 0.0,
|
| 2914 |
+
"step": 4150
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 3.6236933797909407,
|
| 2918 |
+
"grad_norm": NaN,
|
| 2919 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2920 |
+
"loss": 0.0,
|
| 2921 |
+
"step": 4160
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 3.6324041811846692,
|
| 2925 |
+
"grad_norm": NaN,
|
| 2926 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2927 |
+
"loss": 0.0,
|
| 2928 |
+
"step": 4170
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 3.6411149825783973,
|
| 2932 |
+
"grad_norm": NaN,
|
| 2933 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2934 |
+
"loss": 0.0,
|
| 2935 |
+
"step": 4180
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 3.6498257839721253,
|
| 2939 |
+
"grad_norm": NaN,
|
| 2940 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2941 |
+
"loss": 0.0,
|
| 2942 |
+
"step": 4190
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 3.658536585365854,
|
| 2946 |
+
"grad_norm": NaN,
|
| 2947 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2948 |
+
"loss": 0.0,
|
| 2949 |
+
"step": 4200
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 3.667247386759582,
|
| 2953 |
+
"grad_norm": NaN,
|
| 2954 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2955 |
+
"loss": 0.0,
|
| 2956 |
+
"step": 4210
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 3.6759581881533103,
|
| 2960 |
+
"grad_norm": NaN,
|
| 2961 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2962 |
+
"loss": 0.0,
|
| 2963 |
+
"step": 4220
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 3.6846689895470384,
|
| 2967 |
+
"grad_norm": NaN,
|
| 2968 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2969 |
+
"loss": 0.0,
|
| 2970 |
+
"step": 4230
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 3.6933797909407664,
|
| 2974 |
+
"grad_norm": NaN,
|
| 2975 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2976 |
+
"loss": 0.0,
|
| 2977 |
+
"step": 4240
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 3.702090592334495,
|
| 2981 |
+
"grad_norm": NaN,
|
| 2982 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2983 |
+
"loss": 0.0,
|
| 2984 |
+
"step": 4250
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 3.710801393728223,
|
| 2988 |
+
"grad_norm": NaN,
|
| 2989 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2990 |
+
"loss": 0.0,
|
| 2991 |
+
"step": 4260
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 3.7195121951219514,
|
| 2995 |
+
"grad_norm": NaN,
|
| 2996 |
+
"learning_rate": 2.045721369683134e-06,
|
| 2997 |
+
"loss": 0.0,
|
| 2998 |
+
"step": 4270
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 3.7282229965156795,
|
| 3002 |
+
"grad_norm": NaN,
|
| 3003 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3004 |
+
"loss": 0.0,
|
| 3005 |
+
"step": 4280
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 3.7369337979094075,
|
| 3009 |
+
"grad_norm": NaN,
|
| 3010 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3011 |
+
"loss": 0.0,
|
| 3012 |
+
"step": 4290
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 3.745644599303136,
|
| 3016 |
+
"grad_norm": NaN,
|
| 3017 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3018 |
+
"loss": 0.0,
|
| 3019 |
+
"step": 4300
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 3.754355400696864,
|
| 3023 |
+
"grad_norm": NaN,
|
| 3024 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3025 |
+
"loss": 0.0,
|
| 3026 |
+
"step": 4310
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 3.7630662020905925,
|
| 3030 |
+
"grad_norm": NaN,
|
| 3031 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3032 |
+
"loss": 0.0,
|
| 3033 |
+
"step": 4320
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 3.7717770034843205,
|
| 3037 |
+
"grad_norm": NaN,
|
| 3038 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3039 |
+
"loss": 0.0,
|
| 3040 |
+
"step": 4330
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 3.7804878048780486,
|
| 3044 |
+
"grad_norm": NaN,
|
| 3045 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3046 |
+
"loss": 0.0,
|
| 3047 |
+
"step": 4340
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 3.789198606271777,
|
| 3051 |
+
"grad_norm": NaN,
|
| 3052 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3053 |
+
"loss": 0.0,
|
| 3054 |
+
"step": 4350
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 3.797909407665505,
|
| 3058 |
+
"grad_norm": NaN,
|
| 3059 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3060 |
+
"loss": 0.0,
|
| 3061 |
+
"step": 4360
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 3.8066202090592336,
|
| 3065 |
+
"grad_norm": NaN,
|
| 3066 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3067 |
+
"loss": 0.0,
|
| 3068 |
+
"step": 4370
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 3.8153310104529616,
|
| 3072 |
+
"grad_norm": NaN,
|
| 3073 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3074 |
+
"loss": 0.0,
|
| 3075 |
+
"step": 4380
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 3.8240418118466897,
|
| 3079 |
+
"grad_norm": NaN,
|
| 3080 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3081 |
+
"loss": 0.0,
|
| 3082 |
+
"step": 4390
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 3.832752613240418,
|
| 3086 |
+
"grad_norm": NaN,
|
| 3087 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3088 |
+
"loss": 0.0,
|
| 3089 |
+
"step": 4400
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 3.841463414634146,
|
| 3093 |
+
"grad_norm": NaN,
|
| 3094 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3095 |
+
"loss": 0.0,
|
| 3096 |
+
"step": 4410
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 3.8501742160278747,
|
| 3100 |
+
"grad_norm": NaN,
|
| 3101 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3102 |
+
"loss": 0.0,
|
| 3103 |
+
"step": 4420
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 3.8588850174216027,
|
| 3107 |
+
"grad_norm": NaN,
|
| 3108 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3109 |
+
"loss": 0.0,
|
| 3110 |
+
"step": 4430
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 3.8675958188153308,
|
| 3114 |
+
"grad_norm": NaN,
|
| 3115 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3116 |
+
"loss": 0.0,
|
| 3117 |
+
"step": 4440
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 3.8763066202090593,
|
| 3121 |
+
"grad_norm": NaN,
|
| 3122 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3123 |
+
"loss": 0.0,
|
| 3124 |
+
"step": 4450
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 3.8850174216027873,
|
| 3128 |
+
"grad_norm": NaN,
|
| 3129 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3130 |
+
"loss": 0.0,
|
| 3131 |
+
"step": 4460
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 3.8937282229965158,
|
| 3135 |
+
"grad_norm": NaN,
|
| 3136 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3137 |
+
"loss": 0.0,
|
| 3138 |
+
"step": 4470
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 3.902439024390244,
|
| 3142 |
+
"grad_norm": NaN,
|
| 3143 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3144 |
+
"loss": 0.0,
|
| 3145 |
+
"step": 4480
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 3.911149825783972,
|
| 3149 |
+
"grad_norm": NaN,
|
| 3150 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3151 |
+
"loss": 0.0,
|
| 3152 |
+
"step": 4490
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 3.9198606271777003,
|
| 3156 |
+
"grad_norm": NaN,
|
| 3157 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3158 |
+
"loss": 0.0,
|
| 3159 |
+
"step": 4500
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 3.928571428571429,
|
| 3163 |
+
"grad_norm": NaN,
|
| 3164 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3165 |
+
"loss": 0.0,
|
| 3166 |
+
"step": 4510
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 3.937282229965157,
|
| 3170 |
+
"grad_norm": NaN,
|
| 3171 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3172 |
+
"loss": 0.0,
|
| 3173 |
+
"step": 4520
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 3.945993031358885,
|
| 3177 |
+
"grad_norm": NaN,
|
| 3178 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3179 |
+
"loss": 0.0,
|
| 3180 |
+
"step": 4530
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 3.9547038327526134,
|
| 3184 |
+
"grad_norm": NaN,
|
| 3185 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3186 |
+
"loss": 0.0,
|
| 3187 |
+
"step": 4540
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 3.9634146341463414,
|
| 3191 |
+
"grad_norm": NaN,
|
| 3192 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3193 |
+
"loss": 0.0,
|
| 3194 |
+
"step": 4550
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 3.97212543554007,
|
| 3198 |
+
"grad_norm": NaN,
|
| 3199 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3200 |
+
"loss": 0.0,
|
| 3201 |
+
"step": 4560
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 3.980836236933798,
|
| 3205 |
+
"grad_norm": NaN,
|
| 3206 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3207 |
+
"loss": 0.0,
|
| 3208 |
+
"step": 4570
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 3.989547038327526,
|
| 3212 |
+
"grad_norm": NaN,
|
| 3213 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3214 |
+
"loss": 0.0,
|
| 3215 |
+
"step": 4580
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 3.9982578397212545,
|
| 3219 |
+
"grad_norm": NaN,
|
| 3220 |
+
"learning_rate": 2.045721369683134e-06,
|
| 3221 |
+
"loss": 0.0,
|
| 3222 |
+
"step": 4590
|
| 3223 |
+
}
|
| 3224 |
+
],
|
| 3225 |
+
"logging_steps": 10,
|
| 3226 |
+
"max_steps": 5740,
|
| 3227 |
+
"num_input_tokens_seen": 0,
|
| 3228 |
+
"num_train_epochs": 5,
|
| 3229 |
+
"save_steps": 500,
|
| 3230 |
+
"stateful_callbacks": {
|
| 3231 |
+
"TrainerControl": {
|
| 3232 |
+
"args": {
|
| 3233 |
+
"should_epoch_stop": false,
|
| 3234 |
+
"should_evaluate": false,
|
| 3235 |
+
"should_log": false,
|
| 3236 |
+
"should_save": true,
|
| 3237 |
+
"should_training_stop": false
|
| 3238 |
+
},
|
| 3239 |
+
"attributes": {}
|
| 3240 |
+
}
|
| 3241 |
+
},
|
| 3242 |
+
"total_flos": 2263162595155968.0,
|
| 3243 |
+
"train_batch_size": 32,
|
| 3244 |
+
"trial_name": null,
|
| 3245 |
+
"trial_params": null
|
| 3246 |
+
}
|
checkpoint-4592/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:961cda56281c729c0b648a16ab018d2daa38223e675d6788211db692304aaf25
|
| 3 |
+
size 5176
|
checkpoint-4592/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-5740/config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TraXLMistralForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"dropout": 0.1,
|
| 6 |
+
"dynamic_routing": true,
|
| 7 |
+
"ff_expansion_factor": 4,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"id2label": {
|
| 10 |
+
"0": "LABEL_0",
|
| 11 |
+
"1": "LABEL_1",
|
| 12 |
+
"2": "LABEL_2",
|
| 13 |
+
"3": "LABEL_3",
|
| 14 |
+
"4": "LABEL_4"
|
| 15 |
+
},
|
| 16 |
+
"is_decoder": true,
|
| 17 |
+
"label2id": {
|
| 18 |
+
"LABEL_0": 0,
|
| 19 |
+
"LABEL_1": 1,
|
| 20 |
+
"LABEL_2": 2,
|
| 21 |
+
"LABEL_3": 3,
|
| 22 |
+
"LABEL_4": 4
|
| 23 |
+
},
|
| 24 |
+
"max_computation_steps": 5,
|
| 25 |
+
"max_len": 256,
|
| 26 |
+
"memory_size": 256,
|
| 27 |
+
"model_type": "TraXLMistral",
|
| 28 |
+
"n_head": 4,
|
| 29 |
+
"n_layer": 4,
|
| 30 |
+
"rnn_units": 128,
|
| 31 |
+
"sparse_attention": true,
|
| 32 |
+
"torch_dtype": "float32",
|
| 33 |
+
"transformers_version": "4.44.2",
|
| 34 |
+
"vocab_size": 50257
|
| 35 |
+
}
|
checkpoint-5740/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|