add model_bart
Browse files- models/dostoievsky_v1/checkpoint-21500/config.json +71 -0
- models/dostoievsky_v1/checkpoint-21500/generation_config.json +16 -0
- models/dostoievsky_v1/checkpoint-21500/model.safetensors +3 -0
- models/dostoievsky_v1/checkpoint-21500/optimizer.pt +3 -0
- models/dostoievsky_v1/checkpoint-21500/rng_state.pth +3 -0
- models/dostoievsky_v1/checkpoint-21500/scaler.pt +3 -0
- models/dostoievsky_v1/checkpoint-21500/scheduler.pt +3 -0
- models/dostoievsky_v1/checkpoint-21500/tokenizer.json +0 -0
- models/dostoievsky_v1/checkpoint-21500/tokenizer_config.json +16 -0
- models/dostoievsky_v1/checkpoint-21500/trainer_state.json +1547 -0
- models/dostoievsky_v1/checkpoint-21500/training_args.bin +3 -0
- models/dostoievsky_v1/checkpoint-22000/config.json +71 -0
- models/dostoievsky_v1/checkpoint-22000/generation_config.json +16 -0
- models/dostoievsky_v1/checkpoint-22000/model.safetensors +3 -0
- models/dostoievsky_v1/checkpoint-22000/optimizer.pt +3 -0
- models/dostoievsky_v1/checkpoint-22000/rng_state.pth +3 -0
- models/dostoievsky_v1/checkpoint-22000/scaler.pt +3 -0
- models/dostoievsky_v1/checkpoint-22000/scheduler.pt +3 -0
- models/dostoievsky_v1/checkpoint-22000/tokenizer.json +0 -0
- models/dostoievsky_v1/checkpoint-22000/tokenizer_config.json +16 -0
- models/dostoievsky_v1/checkpoint-22000/trainer_state.json +1582 -0
- models/dostoievsky_v1/checkpoint-22000/training_args.bin +3 -0
- models/dostoievsky_v1/checkpoint-22016/config.json +71 -0
- models/dostoievsky_v1/checkpoint-22016/generation_config.json +16 -0
- models/dostoievsky_v1/checkpoint-22016/model.safetensors +3 -0
- models/dostoievsky_v1/checkpoint-22016/optimizer.pt +3 -0
- models/dostoievsky_v1/checkpoint-22016/rng_state.pth +3 -0
- models/dostoievsky_v1/checkpoint-22016/scaler.pt +3 -0
- models/dostoievsky_v1/checkpoint-22016/scheduler.pt +3 -0
- models/dostoievsky_v1/checkpoint-22016/tokenizer.json +0 -0
- models/dostoievsky_v1/checkpoint-22016/tokenizer_config.json +16 -0
- models/dostoievsky_v1/checkpoint-22016/trainer_state.json +1582 -0
- models/dostoievsky_v1/checkpoint-22016/training_args.bin +3 -0
- models/dostoievsky_v1/final_model/config.json +71 -0
- models/dostoievsky_v1/final_model/generation_config.json +16 -0
- models/dostoievsky_v1/final_model/model.safetensors +3 -0
- models/dostoievsky_v1/final_model/tokenizer.json +0 -0
- models/dostoievsky_v1/final_model/tokenizer_config.json +16 -0
- models/dostoievsky_v1/final_model/training_args.bin +3 -0
models/dostoievsky_v1/checkpoint-21500/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.1,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"BartForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.1,
|
| 10 |
+
"bos_token_id": 0,
|
| 11 |
+
"classif_dropout": 0.1,
|
| 12 |
+
"classifier_dropout": 0.0,
|
| 13 |
+
"d_model": 768,
|
| 14 |
+
"decoder_attention_heads": 12,
|
| 15 |
+
"decoder_ffn_dim": 3072,
|
| 16 |
+
"decoder_layerdrop": 0.0,
|
| 17 |
+
"decoder_layers": 6,
|
| 18 |
+
"decoder_start_token_id": 2,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 12,
|
| 22 |
+
"encoder_ffn_dim": 3072,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": true,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1,
|
| 39 |
+
"LABEL_2": 2
|
| 40 |
+
},
|
| 41 |
+
"max_position_embeddings": 1024,
|
| 42 |
+
"model_type": "bart",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": true,
|
| 45 |
+
"pad_token_id": 1,
|
| 46 |
+
"scale_embedding": false,
|
| 47 |
+
"task_specific_params": {
|
| 48 |
+
"summarization": {
|
| 49 |
+
"length_penalty": 1.0,
|
| 50 |
+
"max_length": 128,
|
| 51 |
+
"min_length": 12,
|
| 52 |
+
"num_beams": 4
|
| 53 |
+
},
|
| 54 |
+
"summarization_cnn": {
|
| 55 |
+
"length_penalty": 2.0,
|
| 56 |
+
"max_length": 142,
|
| 57 |
+
"min_length": 56,
|
| 58 |
+
"num_beams": 4
|
| 59 |
+
},
|
| 60 |
+
"summarization_xsum": {
|
| 61 |
+
"length_penalty": 1.0,
|
| 62 |
+
"max_length": 62,
|
| 63 |
+
"min_length": 11,
|
| 64 |
+
"num_beams": 6
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.5.0",
|
| 69 |
+
"use_cache": false,
|
| 70 |
+
"vocab_size": 50265
|
| 71 |
+
}
|
models/dostoievsky_v1/checkpoint-21500/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"early_stopping": true,
|
| 6 |
+
"eos_token_id": [
|
| 7 |
+
2
|
| 8 |
+
],
|
| 9 |
+
"forced_bos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 2,
|
| 11 |
+
"no_repeat_ngram_size": 3,
|
| 12 |
+
"num_beams": 4,
|
| 13 |
+
"pad_token_id": 1,
|
| 14 |
+
"transformers_version": "5.5.0",
|
| 15 |
+
"use_cache": true
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-21500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:acfdbd1127b03f850a3a4fc3107a7df8a0856325e1ce149a5e6d040f6c708422
|
| 3 |
+
size 557912620
|
models/dostoievsky_v1/checkpoint-21500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8bef3b9865106bb639656edae5b6df8906c13a65dc86e92c13bba152cb4e25be
|
| 3 |
+
size 1115583947
|
models/dostoievsky_v1/checkpoint-21500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d9d9a8088d7f25d4d4e3b3b5cf3b6a1457481bb5f05575ed6491980c3ce339a
|
| 3 |
+
size 14645
|
models/dostoievsky_v1/checkpoint-21500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad8834c6e87c2456e1563a68ce36f528eb72a4db9786b554beef824028ea6c9c
|
| 3 |
+
size 1383
|
models/dostoievsky_v1/checkpoint-21500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9528466ddff0be39dea387c74652e8426c9e6284d21cabd291b3a0b686c01d7b
|
| 3 |
+
size 1465
|
models/dostoievsky_v1/checkpoint-21500/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
models/dostoievsky_v1/checkpoint-21500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"mask_token": "<mask>",
|
| 10 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"sep_token": "</s>",
|
| 13 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 14 |
+
"trim_offsets": true,
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-21500/trainer_state.json
ADDED
|
@@ -0,0 +1,1547 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.953125,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 21500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.009084302325581396,
|
| 14 |
+
"grad_norm": 1.9626818895339966,
|
| 15 |
+
"learning_rate": 4.9775163517441866e-05,
|
| 16 |
+
"loss": 3.2140347290039064,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.018168604651162792,
|
| 21 |
+
"grad_norm": 1.511451244354248,
|
| 22 |
+
"learning_rate": 4.9548055959302324e-05,
|
| 23 |
+
"loss": 0.01261173963546753,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.027252906976744186,
|
| 28 |
+
"grad_norm": 0.18194368481636047,
|
| 29 |
+
"learning_rate": 4.9320948401162795e-05,
|
| 30 |
+
"loss": 0.009183254837989808,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.036337209302325583,
|
| 35 |
+
"grad_norm": 0.27592894434928894,
|
| 36 |
+
"learning_rate": 4.909384084302326e-05,
|
| 37 |
+
"loss": 0.0027826914191246034,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.045421511627906974,
|
| 42 |
+
"grad_norm": 1.5584100484848022,
|
| 43 |
+
"learning_rate": 4.8866733284883724e-05,
|
| 44 |
+
"loss": 0.0029672542214393615,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05450581395348837,
|
| 49 |
+
"grad_norm": 0.02698471024632454,
|
| 50 |
+
"learning_rate": 4.863962572674419e-05,
|
| 51 |
+
"loss": 0.004296095669269561,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06359011627906977,
|
| 56 |
+
"grad_norm": 0.137993723154068,
|
| 57 |
+
"learning_rate": 4.841251816860465e-05,
|
| 58 |
+
"loss": 0.0031410756707191466,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07267441860465117,
|
| 63 |
+
"grad_norm": 0.07874622195959091,
|
| 64 |
+
"learning_rate": 4.818541061046512e-05,
|
| 65 |
+
"loss": 0.003037240505218506,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08175872093023256,
|
| 70 |
+
"grad_norm": 0.013660268858075142,
|
| 71 |
+
"learning_rate": 4.795830305232558e-05,
|
| 72 |
+
"loss": 0.0031082597374916076,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.09084302325581395,
|
| 77 |
+
"grad_norm": 0.3354911804199219,
|
| 78 |
+
"learning_rate": 4.7731195494186046e-05,
|
| 79 |
+
"loss": 0.0048285979032516475,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09992732558139535,
|
| 84 |
+
"grad_norm": 0.43667030334472656,
|
| 85 |
+
"learning_rate": 4.750408793604652e-05,
|
| 86 |
+
"loss": 0.0019270157814025878,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10901162790697674,
|
| 91 |
+
"grad_norm": 0.03849343582987785,
|
| 92 |
+
"learning_rate": 4.7276980377906975e-05,
|
| 93 |
+
"loss": 0.001011388897895813,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11809593023255814,
|
| 98 |
+
"grad_norm": 0.017985744401812553,
|
| 99 |
+
"learning_rate": 4.7049872819767446e-05,
|
| 100 |
+
"loss": 0.0018200889229774476,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12718023255813954,
|
| 105 |
+
"grad_norm": 0.05241613835096359,
|
| 106 |
+
"learning_rate": 4.682276526162791e-05,
|
| 107 |
+
"loss": 0.004675151705741882,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13626453488372092,
|
| 112 |
+
"grad_norm": 0.16119782626628876,
|
| 113 |
+
"learning_rate": 4.6595657703488375e-05,
|
| 114 |
+
"loss": 0.0025834646821022034,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.14534883720930233,
|
| 119 |
+
"grad_norm": 0.8985283374786377,
|
| 120 |
+
"learning_rate": 4.636855014534884e-05,
|
| 121 |
+
"loss": 0.001825849711894989,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.15443313953488372,
|
| 126 |
+
"grad_norm": 0.021285999566316605,
|
| 127 |
+
"learning_rate": 4.6141442587209304e-05,
|
| 128 |
+
"loss": 0.004164438843727112,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.16351744186046513,
|
| 133 |
+
"grad_norm": 0.01013926975429058,
|
| 134 |
+
"learning_rate": 4.591433502906977e-05,
|
| 135 |
+
"loss": 0.002582077383995056,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1726017441860465,
|
| 140 |
+
"grad_norm": 0.005713903810828924,
|
| 141 |
+
"learning_rate": 4.568722747093023e-05,
|
| 142 |
+
"loss": 0.0026018735766410827,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1816860465116279,
|
| 147 |
+
"grad_norm": 0.019032707437872887,
|
| 148 |
+
"learning_rate": 4.54601199127907e-05,
|
| 149 |
+
"loss": 0.0017240011692047118,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1907703488372093,
|
| 154 |
+
"grad_norm": 0.03191379830241203,
|
| 155 |
+
"learning_rate": 4.523301235465117e-05,
|
| 156 |
+
"loss": 0.0022749193012714386,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1998546511627907,
|
| 161 |
+
"grad_norm": 0.12945719063282013,
|
| 162 |
+
"learning_rate": 4.5005904796511626e-05,
|
| 163 |
+
"loss": 0.0017636509239673615,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2089389534883721,
|
| 168 |
+
"grad_norm": 0.01494428887963295,
|
| 169 |
+
"learning_rate": 4.47787972383721e-05,
|
| 170 |
+
"loss": 0.0017163331806659698,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2180232558139535,
|
| 175 |
+
"grad_norm": 0.08089974522590637,
|
| 176 |
+
"learning_rate": 4.455168968023256e-05,
|
| 177 |
+
"loss": 0.0023377402126789092,
|
| 178 |
+
"step": 2400
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.22710755813953487,
|
| 182 |
+
"grad_norm": 0.004136559087783098,
|
| 183 |
+
"learning_rate": 4.4324582122093026e-05,
|
| 184 |
+
"loss": 0.001353910267353058,
|
| 185 |
+
"step": 2500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.23619186046511628,
|
| 189 |
+
"grad_norm": 0.25111478567123413,
|
| 190 |
+
"learning_rate": 4.409747456395349e-05,
|
| 191 |
+
"loss": 0.0022939679026603697,
|
| 192 |
+
"step": 2600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.24527616279069767,
|
| 196 |
+
"grad_norm": 0.963623583316803,
|
| 197 |
+
"learning_rate": 4.3870367005813955e-05,
|
| 198 |
+
"loss": 0.0018738456070423125,
|
| 199 |
+
"step": 2700
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2543604651162791,
|
| 203 |
+
"grad_norm": 0.6530899405479431,
|
| 204 |
+
"learning_rate": 4.364325944767442e-05,
|
| 205 |
+
"loss": 0.0015843257308006287,
|
| 206 |
+
"step": 2800
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.26344476744186046,
|
| 210 |
+
"grad_norm": 0.0018010369967669249,
|
| 211 |
+
"learning_rate": 4.3416151889534884e-05,
|
| 212 |
+
"loss": 0.0006610354781150817,
|
| 213 |
+
"step": 2900
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.27252906976744184,
|
| 217 |
+
"grad_norm": 0.1644030064344406,
|
| 218 |
+
"learning_rate": 4.318904433139535e-05,
|
| 219 |
+
"loss": 0.0031197205185890197,
|
| 220 |
+
"step": 3000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.28161337209302323,
|
| 224 |
+
"grad_norm": 0.14440134167671204,
|
| 225 |
+
"learning_rate": 4.296193677325582e-05,
|
| 226 |
+
"loss": 0.0011694706976413728,
|
| 227 |
+
"step": 3100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.29069767441860467,
|
| 231 |
+
"grad_norm": 0.0124755734577775,
|
| 232 |
+
"learning_rate": 4.273482921511628e-05,
|
| 233 |
+
"loss": 0.0013250903785228729,
|
| 234 |
+
"step": 3200
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.29978197674418605,
|
| 238 |
+
"grad_norm": 0.15910762548446655,
|
| 239 |
+
"learning_rate": 4.250772165697675e-05,
|
| 240 |
+
"loss": 0.0007842753082513809,
|
| 241 |
+
"step": 3300
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.30886627906976744,
|
| 245 |
+
"grad_norm": 0.011540662497282028,
|
| 246 |
+
"learning_rate": 4.228061409883721e-05,
|
| 247 |
+
"loss": 0.0018773898482322693,
|
| 248 |
+
"step": 3400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3179505813953488,
|
| 252 |
+
"grad_norm": 0.005148735363036394,
|
| 253 |
+
"learning_rate": 4.205350654069768e-05,
|
| 254 |
+
"loss": 0.001056775525212288,
|
| 255 |
+
"step": 3500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.32703488372093026,
|
| 259 |
+
"grad_norm": 0.007442768197506666,
|
| 260 |
+
"learning_rate": 4.182639898255814e-05,
|
| 261 |
+
"loss": 0.0014348265528678894,
|
| 262 |
+
"step": 3600
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.33611918604651164,
|
| 266 |
+
"grad_norm": 0.00586783979088068,
|
| 267 |
+
"learning_rate": 4.1599291424418606e-05,
|
| 268 |
+
"loss": 0.0014780126512050629,
|
| 269 |
+
"step": 3700
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.345203488372093,
|
| 273 |
+
"grad_norm": 0.006909696385264397,
|
| 274 |
+
"learning_rate": 4.137218386627907e-05,
|
| 275 |
+
"loss": 0.0019674110412597656,
|
| 276 |
+
"step": 3800
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3542877906976744,
|
| 280 |
+
"grad_norm": 0.020171863958239555,
|
| 281 |
+
"learning_rate": 4.1145076308139535e-05,
|
| 282 |
+
"loss": 0.00249418705701828,
|
| 283 |
+
"step": 3900
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3633720930232558,
|
| 287 |
+
"grad_norm": 0.031723715364933014,
|
| 288 |
+
"learning_rate": 4.091796875e-05,
|
| 289 |
+
"loss": 0.0030515387654304503,
|
| 290 |
+
"step": 4000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.37245639534883723,
|
| 294 |
+
"grad_norm": 0.30729565024375916,
|
| 295 |
+
"learning_rate": 4.069086119186047e-05,
|
| 296 |
+
"loss": 0.0009509802609682083,
|
| 297 |
+
"step": 4100
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3815406976744186,
|
| 301 |
+
"grad_norm": 0.0021714100148528814,
|
| 302 |
+
"learning_rate": 4.046375363372093e-05,
|
| 303 |
+
"loss": 0.0013920699059963226,
|
| 304 |
+
"step": 4200
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.390625,
|
| 308 |
+
"grad_norm": 0.001017636270262301,
|
| 309 |
+
"learning_rate": 4.02366460755814e-05,
|
| 310 |
+
"loss": 0.0010310819000005722,
|
| 311 |
+
"step": 4300
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3997093023255814,
|
| 315 |
+
"grad_norm": 0.0009663284290581942,
|
| 316 |
+
"learning_rate": 4.0009538517441864e-05,
|
| 317 |
+
"loss": 0.001098722442984581,
|
| 318 |
+
"step": 4400
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.40879360465116277,
|
| 322 |
+
"grad_norm": 0.003507931251078844,
|
| 323 |
+
"learning_rate": 3.978243095930232e-05,
|
| 324 |
+
"loss": 0.0013840131461620331,
|
| 325 |
+
"step": 4500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4178779069767442,
|
| 329 |
+
"grad_norm": 0.21988199651241302,
|
| 330 |
+
"learning_rate": 3.955532340116279e-05,
|
| 331 |
+
"loss": 0.0017136169970035553,
|
| 332 |
+
"step": 4600
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4269622093023256,
|
| 336 |
+
"grad_norm": 0.0012972657568752766,
|
| 337 |
+
"learning_rate": 3.932821584302326e-05,
|
| 338 |
+
"loss": 0.0008040449023246765,
|
| 339 |
+
"step": 4700
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.436046511627907,
|
| 343 |
+
"grad_norm": 0.033857911825180054,
|
| 344 |
+
"learning_rate": 3.910110828488372e-05,
|
| 345 |
+
"loss": 0.0005662231892347336,
|
| 346 |
+
"step": 4800
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.44513081395348836,
|
| 350 |
+
"grad_norm": 0.0008128180052153766,
|
| 351 |
+
"learning_rate": 3.8874000726744187e-05,
|
| 352 |
+
"loss": 0.001040520742535591,
|
| 353 |
+
"step": 4900
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.45421511627906974,
|
| 357 |
+
"grad_norm": 0.0017111338675022125,
|
| 358 |
+
"learning_rate": 3.864689316860465e-05,
|
| 359 |
+
"loss": 0.0010922805964946748,
|
| 360 |
+
"step": 5000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.4632994186046512,
|
| 364 |
+
"grad_norm": 0.0013291583163663745,
|
| 365 |
+
"learning_rate": 3.841978561046512e-05,
|
| 366 |
+
"loss": 0.0005277743935585021,
|
| 367 |
+
"step": 5100
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.47238372093023256,
|
| 371 |
+
"grad_norm": 0.09846807271242142,
|
| 372 |
+
"learning_rate": 3.819267805232558e-05,
|
| 373 |
+
"loss": 0.0010997675359249114,
|
| 374 |
+
"step": 5200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.48146802325581395,
|
| 378 |
+
"grad_norm": 0.005185174290090799,
|
| 379 |
+
"learning_rate": 3.796557049418605e-05,
|
| 380 |
+
"loss": 0.000698111355304718,
|
| 381 |
+
"step": 5300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.49055232558139533,
|
| 385 |
+
"grad_norm": 0.019086388871073723,
|
| 386 |
+
"learning_rate": 3.7738462936046515e-05,
|
| 387 |
+
"loss": 0.0009016367793083191,
|
| 388 |
+
"step": 5400
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.49963662790697677,
|
| 392 |
+
"grad_norm": 0.05122831463813782,
|
| 393 |
+
"learning_rate": 3.751135537790697e-05,
|
| 394 |
+
"loss": 0.0015524370968341826,
|
| 395 |
+
"step": 5500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5087209302325582,
|
| 399 |
+
"grad_norm": 0.012596211396157742,
|
| 400 |
+
"learning_rate": 3.7284247819767444e-05,
|
| 401 |
+
"loss": 0.00220917209982872,
|
| 402 |
+
"step": 5600
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5178052325581395,
|
| 406 |
+
"grad_norm": 0.009361029602587223,
|
| 407 |
+
"learning_rate": 3.705714026162791e-05,
|
| 408 |
+
"loss": 0.0010362663865089417,
|
| 409 |
+
"step": 5700
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5268895348837209,
|
| 413 |
+
"grad_norm": 0.5205378532409668,
|
| 414 |
+
"learning_rate": 3.683003270348837e-05,
|
| 415 |
+
"loss": 0.0022550773620605468,
|
| 416 |
+
"step": 5800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5359738372093024,
|
| 420 |
+
"grad_norm": 0.0006470124353654683,
|
| 421 |
+
"learning_rate": 3.660292514534884e-05,
|
| 422 |
+
"loss": 0.0011264414340257645,
|
| 423 |
+
"step": 5900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5450581395348837,
|
| 427 |
+
"grad_norm": 0.058231666684150696,
|
| 428 |
+
"learning_rate": 3.63758175872093e-05,
|
| 429 |
+
"loss": 0.0011266635358333588,
|
| 430 |
+
"step": 6000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5541424418604651,
|
| 434 |
+
"grad_norm": 0.0013441125629469752,
|
| 435 |
+
"learning_rate": 3.614871002906977e-05,
|
| 436 |
+
"loss": 0.0008001205325126648,
|
| 437 |
+
"step": 6100
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5632267441860465,
|
| 441 |
+
"grad_norm": 0.0021539030130952597,
|
| 442 |
+
"learning_rate": 3.592160247093023e-05,
|
| 443 |
+
"loss": 0.0008849448710680007,
|
| 444 |
+
"step": 6200
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5723110465116279,
|
| 448 |
+
"grad_norm": 0.9019960761070251,
|
| 449 |
+
"learning_rate": 3.56944949127907e-05,
|
| 450 |
+
"loss": 0.0006517694145441055,
|
| 451 |
+
"step": 6300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5813953488372093,
|
| 455 |
+
"grad_norm": 0.001283760997466743,
|
| 456 |
+
"learning_rate": 3.546738735465117e-05,
|
| 457 |
+
"loss": 0.0021727623045444486,
|
| 458 |
+
"step": 6400
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5904796511627907,
|
| 462 |
+
"grad_norm": 0.005666423588991165,
|
| 463 |
+
"learning_rate": 3.5240279796511624e-05,
|
| 464 |
+
"loss": 0.0012587083876132966,
|
| 465 |
+
"step": 6500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5995639534883721,
|
| 469 |
+
"grad_norm": 1.4842944145202637,
|
| 470 |
+
"learning_rate": 3.5013172238372096e-05,
|
| 471 |
+
"loss": 0.001180112287402153,
|
| 472 |
+
"step": 6600
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6086482558139535,
|
| 476 |
+
"grad_norm": 0.02973734401166439,
|
| 477 |
+
"learning_rate": 3.478606468023256e-05,
|
| 478 |
+
"loss": 0.0010163599252700805,
|
| 479 |
+
"step": 6700
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6177325581395349,
|
| 483 |
+
"grad_norm": 0.00399527233093977,
|
| 484 |
+
"learning_rate": 3.4558957122093024e-05,
|
| 485 |
+
"loss": 0.0010414445400238038,
|
| 486 |
+
"step": 6800
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6268168604651163,
|
| 490 |
+
"grad_norm": 0.39672717452049255,
|
| 491 |
+
"learning_rate": 3.433184956395349e-05,
|
| 492 |
+
"loss": 0.0012094499170780182,
|
| 493 |
+
"step": 6900
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6359011627906976,
|
| 497 |
+
"grad_norm": 0.5308993458747864,
|
| 498 |
+
"learning_rate": 3.410474200581395e-05,
|
| 499 |
+
"loss": 0.0006021633744239807,
|
| 500 |
+
"step": 7000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6449854651162791,
|
| 504 |
+
"grad_norm": 0.003343795659020543,
|
| 505 |
+
"learning_rate": 3.3877634447674425e-05,
|
| 506 |
+
"loss": 0.0005116893351078033,
|
| 507 |
+
"step": 7100
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6540697674418605,
|
| 511 |
+
"grad_norm": 0.0033038391266018152,
|
| 512 |
+
"learning_rate": 3.365052688953488e-05,
|
| 513 |
+
"loss": 0.002106922417879105,
|
| 514 |
+
"step": 7200
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6631540697674418,
|
| 518 |
+
"grad_norm": 0.004004980903118849,
|
| 519 |
+
"learning_rate": 3.3423419331395353e-05,
|
| 520 |
+
"loss": 0.0010233993828296661,
|
| 521 |
+
"step": 7300
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6722383720930233,
|
| 525 |
+
"grad_norm": 0.00490298168733716,
|
| 526 |
+
"learning_rate": 3.319631177325582e-05,
|
| 527 |
+
"loss": 0.0007062336057424546,
|
| 528 |
+
"step": 7400
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6813226744186046,
|
| 532 |
+
"grad_norm": 0.04243200644850731,
|
| 533 |
+
"learning_rate": 3.2969204215116276e-05,
|
| 534 |
+
"loss": 0.0005479569733142852,
|
| 535 |
+
"step": 7500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.690406976744186,
|
| 539 |
+
"grad_norm": 0.008556894026696682,
|
| 540 |
+
"learning_rate": 3.274209665697675e-05,
|
| 541 |
+
"loss": 0.0010938134789466858,
|
| 542 |
+
"step": 7600
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6994912790697675,
|
| 546 |
+
"grad_norm": 0.016938723623752594,
|
| 547 |
+
"learning_rate": 3.251498909883721e-05,
|
| 548 |
+
"loss": 0.0009466408193111419,
|
| 549 |
+
"step": 7700
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7085755813953488,
|
| 553 |
+
"grad_norm": 0.09418574720621109,
|
| 554 |
+
"learning_rate": 3.2287881540697676e-05,
|
| 555 |
+
"loss": 0.00145876482129097,
|
| 556 |
+
"step": 7800
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7176598837209303,
|
| 560 |
+
"grad_norm": 0.002359782112762332,
|
| 561 |
+
"learning_rate": 3.206077398255814e-05,
|
| 562 |
+
"loss": 0.0015742655098438263,
|
| 563 |
+
"step": 7900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7267441860465116,
|
| 567 |
+
"grad_norm": 0.3374776840209961,
|
| 568 |
+
"learning_rate": 3.1833666424418605e-05,
|
| 569 |
+
"loss": 0.0007466593384742737,
|
| 570 |
+
"step": 8000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.735828488372093,
|
| 574 |
+
"grad_norm": 0.000735403154976666,
|
| 575 |
+
"learning_rate": 3.1606558866279076e-05,
|
| 576 |
+
"loss": 0.0008171546459197998,
|
| 577 |
+
"step": 8100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7449127906976745,
|
| 581 |
+
"grad_norm": 0.01963644102215767,
|
| 582 |
+
"learning_rate": 3.1379451308139533e-05,
|
| 583 |
+
"loss": 0.0007363802939653396,
|
| 584 |
+
"step": 8200
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7539970930232558,
|
| 588 |
+
"grad_norm": 0.09964141249656677,
|
| 589 |
+
"learning_rate": 3.1152343750000005e-05,
|
| 590 |
+
"loss": 0.00030826406553387644,
|
| 591 |
+
"step": 8300
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7630813953488372,
|
| 595 |
+
"grad_norm": 0.0029934593476355076,
|
| 596 |
+
"learning_rate": 3.092523619186047e-05,
|
| 597 |
+
"loss": 0.0003284827247262001,
|
| 598 |
+
"step": 8400
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7721656976744186,
|
| 602 |
+
"grad_norm": 0.002162993187084794,
|
| 603 |
+
"learning_rate": 3.069812863372093e-05,
|
| 604 |
+
"loss": 0.0010288888961076737,
|
| 605 |
+
"step": 8500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.78125,
|
| 609 |
+
"grad_norm": 0.013634397648274899,
|
| 610 |
+
"learning_rate": 3.0471021075581398e-05,
|
| 611 |
+
"loss": 0.0006983913481235504,
|
| 612 |
+
"step": 8600
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7903343023255814,
|
| 616 |
+
"grad_norm": 0.0833001658320427,
|
| 617 |
+
"learning_rate": 3.0243913517441862e-05,
|
| 618 |
+
"loss": 0.0009967025369405746,
|
| 619 |
+
"step": 8700
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7994186046511628,
|
| 623 |
+
"grad_norm": 0.4697379767894745,
|
| 624 |
+
"learning_rate": 3.0016805959302323e-05,
|
| 625 |
+
"loss": 0.0007716407626867294,
|
| 626 |
+
"step": 8800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8085029069767442,
|
| 630 |
+
"grad_norm": 0.04588587209582329,
|
| 631 |
+
"learning_rate": 2.978969840116279e-05,
|
| 632 |
+
"loss": 0.000263803955167532,
|
| 633 |
+
"step": 8900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8175872093023255,
|
| 637 |
+
"grad_norm": 0.003420337103307247,
|
| 638 |
+
"learning_rate": 2.956259084302326e-05,
|
| 639 |
+
"loss": 0.0009599439054727555,
|
| 640 |
+
"step": 9000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.826671511627907,
|
| 644 |
+
"grad_norm": 0.0024436134845018387,
|
| 645 |
+
"learning_rate": 2.9335483284883724e-05,
|
| 646 |
+
"loss": 0.012270723581314086,
|
| 647 |
+
"step": 9100
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8357558139534884,
|
| 651 |
+
"grad_norm": 0.001418731757439673,
|
| 652 |
+
"learning_rate": 2.9108375726744185e-05,
|
| 653 |
+
"loss": 0.0009713788330554962,
|
| 654 |
+
"step": 9200
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8448401162790697,
|
| 658 |
+
"grad_norm": 0.0026375153101980686,
|
| 659 |
+
"learning_rate": 2.8881268168604652e-05,
|
| 660 |
+
"loss": 0.0007654589414596558,
|
| 661 |
+
"step": 9300
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8539244186046512,
|
| 665 |
+
"grad_norm": 0.0013760777655988932,
|
| 666 |
+
"learning_rate": 2.865416061046512e-05,
|
| 667 |
+
"loss": 0.0006276721507310867,
|
| 668 |
+
"step": 9400
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8630087209302325,
|
| 672 |
+
"grad_norm": 0.0033582421019673347,
|
| 673 |
+
"learning_rate": 2.842705305232558e-05,
|
| 674 |
+
"loss": 0.0003520375117659569,
|
| 675 |
+
"step": 9500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.872093023255814,
|
| 679 |
+
"grad_norm": 0.0043012769892811775,
|
| 680 |
+
"learning_rate": 2.819994549418605e-05,
|
| 681 |
+
"loss": 0.0011182524263858794,
|
| 682 |
+
"step": 9600
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8811773255813954,
|
| 686 |
+
"grad_norm": 0.0015025343745946884,
|
| 687 |
+
"learning_rate": 2.7972837936046514e-05,
|
| 688 |
+
"loss": 0.00044553544372320173,
|
| 689 |
+
"step": 9700
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8902616279069767,
|
| 693 |
+
"grad_norm": 0.0018241156358271837,
|
| 694 |
+
"learning_rate": 2.7745730377906975e-05,
|
| 695 |
+
"loss": 0.00023134740069508553,
|
| 696 |
+
"step": 9800
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8993459302325582,
|
| 700 |
+
"grad_norm": 0.001082456554286182,
|
| 701 |
+
"learning_rate": 2.7518622819767442e-05,
|
| 702 |
+
"loss": 0.001018296480178833,
|
| 703 |
+
"step": 9900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.9084302325581395,
|
| 707 |
+
"grad_norm": 0.01155087724328041,
|
| 708 |
+
"learning_rate": 2.729151526162791e-05,
|
| 709 |
+
"loss": 0.0005589094385504722,
|
| 710 |
+
"step": 10000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.9175145348837209,
|
| 714 |
+
"grad_norm": 0.0022482366766780615,
|
| 715 |
+
"learning_rate": 2.7064407703488375e-05,
|
| 716 |
+
"loss": 0.000512191392481327,
|
| 717 |
+
"step": 10100
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9265988372093024,
|
| 721 |
+
"grad_norm": 0.20549768209457397,
|
| 722 |
+
"learning_rate": 2.6837300145348836e-05,
|
| 723 |
+
"loss": 0.00038118865340948106,
|
| 724 |
+
"step": 10200
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9356831395348837,
|
| 728 |
+
"grad_norm": 0.0013188497396185994,
|
| 729 |
+
"learning_rate": 2.6610192587209304e-05,
|
| 730 |
+
"loss": 0.0003574254736304283,
|
| 731 |
+
"step": 10300
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9447674418604651,
|
| 735 |
+
"grad_norm": 0.0006103936466388404,
|
| 736 |
+
"learning_rate": 2.638308502906977e-05,
|
| 737 |
+
"loss": 0.00024356411769986154,
|
| 738 |
+
"step": 10400
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9538517441860465,
|
| 742 |
+
"grad_norm": 0.00047453015577048063,
|
| 743 |
+
"learning_rate": 2.6155977470930233e-05,
|
| 744 |
+
"loss": 0.0013259868323802948,
|
| 745 |
+
"step": 10500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9629360465116279,
|
| 749 |
+
"grad_norm": 0.025760261341929436,
|
| 750 |
+
"learning_rate": 2.59288699127907e-05,
|
| 751 |
+
"loss": 0.0006390263140201569,
|
| 752 |
+
"step": 10600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9720203488372093,
|
| 756 |
+
"grad_norm": 0.0025282115675508976,
|
| 757 |
+
"learning_rate": 2.5701762354651165e-05,
|
| 758 |
+
"loss": 0.001759422868490219,
|
| 759 |
+
"step": 10700
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.9811046511627907,
|
| 763 |
+
"grad_norm": 2.990682601928711,
|
| 764 |
+
"learning_rate": 2.5474654796511626e-05,
|
| 765 |
+
"loss": 0.00041438560932874677,
|
| 766 |
+
"step": 10800
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9901889534883721,
|
| 770 |
+
"grad_norm": 0.0036354295443743467,
|
| 771 |
+
"learning_rate": 2.5247547238372094e-05,
|
| 772 |
+
"loss": 0.000467718243598938,
|
| 773 |
+
"step": 10900
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9992732558139535,
|
| 777 |
+
"grad_norm": 0.02648981101810932,
|
| 778 |
+
"learning_rate": 2.502043968023256e-05,
|
| 779 |
+
"loss": 0.0007214382290840149,
|
| 780 |
+
"step": 11000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0,
|
| 784 |
+
"eval_loss": 8.017317304620519e-05,
|
| 785 |
+
"eval_runtime": 132.6261,
|
| 786 |
+
"eval_samples_per_second": 165.993,
|
| 787 |
+
"eval_steps_per_second": 20.75,
|
| 788 |
+
"step": 11008
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.008357558139535,
|
| 792 |
+
"grad_norm": 0.0011333210859447718,
|
| 793 |
+
"learning_rate": 2.4793332122093026e-05,
|
| 794 |
+
"loss": 0.000609114095568657,
|
| 795 |
+
"step": 11100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.0174418604651163,
|
| 799 |
+
"grad_norm": 0.0006669509457424283,
|
| 800 |
+
"learning_rate": 2.4566224563953487e-05,
|
| 801 |
+
"loss": 0.000269341841340065,
|
| 802 |
+
"step": 11200
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.0265261627906976,
|
| 806 |
+
"grad_norm": 0.0006085751811042428,
|
| 807 |
+
"learning_rate": 2.4339117005813955e-05,
|
| 808 |
+
"loss": 0.00018008088693022728,
|
| 809 |
+
"step": 11300
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.035610465116279,
|
| 813 |
+
"grad_norm": 0.00029874974279664457,
|
| 814 |
+
"learning_rate": 2.411200944767442e-05,
|
| 815 |
+
"loss": 8.143479935824871e-05,
|
| 816 |
+
"step": 11400
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.0446947674418605,
|
| 820 |
+
"grad_norm": 0.0020065035205334425,
|
| 821 |
+
"learning_rate": 2.3884901889534887e-05,
|
| 822 |
+
"loss": 0.0004405264183878899,
|
| 823 |
+
"step": 11500
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.0537790697674418,
|
| 827 |
+
"grad_norm": 0.3139957785606384,
|
| 828 |
+
"learning_rate": 2.3657794331395348e-05,
|
| 829 |
+
"loss": 0.0007688279449939728,
|
| 830 |
+
"step": 11600
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.0628633720930232,
|
| 834 |
+
"grad_norm": 0.014869201928377151,
|
| 835 |
+
"learning_rate": 2.3430686773255813e-05,
|
| 836 |
+
"loss": 0.00045145414769649507,
|
| 837 |
+
"step": 11700
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.0719476744186047,
|
| 841 |
+
"grad_norm": 0.03592238947749138,
|
| 842 |
+
"learning_rate": 2.320357921511628e-05,
|
| 843 |
+
"loss": 0.0013660797476768493,
|
| 844 |
+
"step": 11800
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.081031976744186,
|
| 848 |
+
"grad_norm": 0.010270297527313232,
|
| 849 |
+
"learning_rate": 2.2976471656976745e-05,
|
| 850 |
+
"loss": 0.0005171676725149155,
|
| 851 |
+
"step": 11900
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.0901162790697674,
|
| 855 |
+
"grad_norm": 0.0026885548140853643,
|
| 856 |
+
"learning_rate": 2.2749364098837213e-05,
|
| 857 |
+
"loss": 0.00046057451516389846,
|
| 858 |
+
"step": 12000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.099200581395349,
|
| 862 |
+
"grad_norm": 0.005681134294718504,
|
| 863 |
+
"learning_rate": 2.2522256540697674e-05,
|
| 864 |
+
"loss": 0.0006235280632972718,
|
| 865 |
+
"step": 12100
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.1082848837209303,
|
| 869 |
+
"grad_norm": 0.0028001824393868446,
|
| 870 |
+
"learning_rate": 2.2295148982558138e-05,
|
| 871 |
+
"loss": 0.0008198145776987075,
|
| 872 |
+
"step": 12200
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.1173691860465116,
|
| 876 |
+
"grad_norm": 0.0006385694723576307,
|
| 877 |
+
"learning_rate": 2.2068041424418606e-05,
|
| 878 |
+
"loss": 0.0003722207620739937,
|
| 879 |
+
"step": 12300
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.1264534883720931,
|
| 883 |
+
"grad_norm": 0.0008649047813378274,
|
| 884 |
+
"learning_rate": 2.184093386627907e-05,
|
| 885 |
+
"loss": 0.0003394853696227074,
|
| 886 |
+
"step": 12400
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.1355377906976745,
|
| 890 |
+
"grad_norm": 0.004796006251126528,
|
| 891 |
+
"learning_rate": 2.1613826308139538e-05,
|
| 892 |
+
"loss": 0.00042502928525209424,
|
| 893 |
+
"step": 12500
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.1446220930232558,
|
| 897 |
+
"grad_norm": 0.001498043886385858,
|
| 898 |
+
"learning_rate": 2.138671875e-05,
|
| 899 |
+
"loss": 0.0009064202010631562,
|
| 900 |
+
"step": 12600
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.1537063953488371,
|
| 904 |
+
"grad_norm": 0.01664622500538826,
|
| 905 |
+
"learning_rate": 2.1159611191860464e-05,
|
| 906 |
+
"loss": 0.0002878718450665474,
|
| 907 |
+
"step": 12700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.1627906976744187,
|
| 911 |
+
"grad_norm": 0.010639426298439503,
|
| 912 |
+
"learning_rate": 2.093250363372093e-05,
|
| 913 |
+
"loss": 0.0011860337108373642,
|
| 914 |
+
"step": 12800
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.171875,
|
| 918 |
+
"grad_norm": 0.050644513219594955,
|
| 919 |
+
"learning_rate": 2.0705396075581396e-05,
|
| 920 |
+
"loss": 0.0006260576844215393,
|
| 921 |
+
"step": 12900
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.1809593023255813,
|
| 925 |
+
"grad_norm": 0.001398528809659183,
|
| 926 |
+
"learning_rate": 2.0478288517441864e-05,
|
| 927 |
+
"loss": 0.0001833665184676647,
|
| 928 |
+
"step": 13000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.1900436046511629,
|
| 932 |
+
"grad_norm": 0.0057320562191307545,
|
| 933 |
+
"learning_rate": 2.0251180959302325e-05,
|
| 934 |
+
"loss": 0.00032943256199359896,
|
| 935 |
+
"step": 13100
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.1991279069767442,
|
| 939 |
+
"grad_norm": 0.0002997924748342484,
|
| 940 |
+
"learning_rate": 2.002407340116279e-05,
|
| 941 |
+
"loss": 6.249572150409222e-05,
|
| 942 |
+
"step": 13200
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.2082122093023255,
|
| 946 |
+
"grad_norm": 0.0008130258647724986,
|
| 947 |
+
"learning_rate": 1.9796965843023257e-05,
|
| 948 |
+
"loss": 0.00026577839627861977,
|
| 949 |
+
"step": 13300
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.2172965116279069,
|
| 953 |
+
"grad_norm": 0.0006040658336132765,
|
| 954 |
+
"learning_rate": 1.956985828488372e-05,
|
| 955 |
+
"loss": 0.00023248378187417984,
|
| 956 |
+
"step": 13400
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.2263808139534884,
|
| 960 |
+
"grad_norm": 0.000803643895778805,
|
| 961 |
+
"learning_rate": 1.934275072674419e-05,
|
| 962 |
+
"loss": 0.0002730824239552021,
|
| 963 |
+
"step": 13500
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.2354651162790697,
|
| 967 |
+
"grad_norm": 0.0005482266424223781,
|
| 968 |
+
"learning_rate": 1.911564316860465e-05,
|
| 969 |
+
"loss": 0.0005848826467990875,
|
| 970 |
+
"step": 13600
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.244549418604651,
|
| 974 |
+
"grad_norm": 0.0034759771078824997,
|
| 975 |
+
"learning_rate": 1.8888535610465115e-05,
|
| 976 |
+
"loss": 0.0005462893471121788,
|
| 977 |
+
"step": 13700
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.2536337209302326,
|
| 981 |
+
"grad_norm": 0.0002226918877568096,
|
| 982 |
+
"learning_rate": 1.8661428052325583e-05,
|
| 983 |
+
"loss": 7.945694960653782e-05,
|
| 984 |
+
"step": 13800
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.262718023255814,
|
| 988 |
+
"grad_norm": 0.002374310279265046,
|
| 989 |
+
"learning_rate": 1.8434320494186047e-05,
|
| 990 |
+
"loss": 0.000744214728474617,
|
| 991 |
+
"step": 13900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.2718023255813953,
|
| 995 |
+
"grad_norm": 0.0009076696587726474,
|
| 996 |
+
"learning_rate": 1.8207212936046515e-05,
|
| 997 |
+
"loss": 0.0003250580281019211,
|
| 998 |
+
"step": 14000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.2808866279069768,
|
| 1002 |
+
"grad_norm": 0.010181100107729435,
|
| 1003 |
+
"learning_rate": 1.7980105377906976e-05,
|
| 1004 |
+
"loss": 0.0005765938758850097,
|
| 1005 |
+
"step": 14100
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.2899709302325582,
|
| 1009 |
+
"grad_norm": 0.0012442917795851827,
|
| 1010 |
+
"learning_rate": 1.775299781976744e-05,
|
| 1011 |
+
"loss": 0.00032320961356163024,
|
| 1012 |
+
"step": 14200
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.2990552325581395,
|
| 1016 |
+
"grad_norm": 0.0005617383285425603,
|
| 1017 |
+
"learning_rate": 1.752589026162791e-05,
|
| 1018 |
+
"loss": 0.0001776321791112423,
|
| 1019 |
+
"step": 14300
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.308139534883721,
|
| 1023 |
+
"grad_norm": 0.0010064981179311872,
|
| 1024 |
+
"learning_rate": 1.7298782703488373e-05,
|
| 1025 |
+
"loss": 0.00012774170376360415,
|
| 1026 |
+
"step": 14400
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.3172238372093024,
|
| 1030 |
+
"grad_norm": 0.00039220438338816166,
|
| 1031 |
+
"learning_rate": 1.707167514534884e-05,
|
| 1032 |
+
"loss": 0.0006190959364175796,
|
| 1033 |
+
"step": 14500
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.3263081395348837,
|
| 1037 |
+
"grad_norm": 0.0009137202869169414,
|
| 1038 |
+
"learning_rate": 1.6844567587209302e-05,
|
| 1039 |
+
"loss": 0.0007051125913858414,
|
| 1040 |
+
"step": 14600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.3353924418604652,
|
| 1044 |
+
"grad_norm": 0.0006007241318002343,
|
| 1045 |
+
"learning_rate": 1.6617460029069766e-05,
|
| 1046 |
+
"loss": 7.074063178151846e-05,
|
| 1047 |
+
"step": 14700
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.3444767441860466,
|
| 1051 |
+
"grad_norm": 0.0006039150175638497,
|
| 1052 |
+
"learning_rate": 1.6390352470930234e-05,
|
| 1053 |
+
"loss": 0.000489293597638607,
|
| 1054 |
+
"step": 14800
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.353561046511628,
|
| 1058 |
+
"grad_norm": 0.0014955669175833464,
|
| 1059 |
+
"learning_rate": 1.61632449127907e-05,
|
| 1060 |
+
"loss": 0.0004155828058719635,
|
| 1061 |
+
"step": 14900
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.3626453488372092,
|
| 1065 |
+
"grad_norm": 0.1508745402097702,
|
| 1066 |
+
"learning_rate": 1.5936137354651163e-05,
|
| 1067 |
+
"loss": 0.00024092141538858413,
|
| 1068 |
+
"step": 15000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.3717296511627908,
|
| 1072 |
+
"grad_norm": 0.0008113393560051918,
|
| 1073 |
+
"learning_rate": 1.5709029796511627e-05,
|
| 1074 |
+
"loss": 0.00040562458336353304,
|
| 1075 |
+
"step": 15100
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.380813953488372,
|
| 1079 |
+
"grad_norm": 0.0026459796354174614,
|
| 1080 |
+
"learning_rate": 1.5481922238372092e-05,
|
| 1081 |
+
"loss": 0.0007934534549713135,
|
| 1082 |
+
"step": 15200
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.3898982558139534,
|
| 1086 |
+
"grad_norm": 0.0005005365237593651,
|
| 1087 |
+
"learning_rate": 1.525481468023256e-05,
|
| 1088 |
+
"loss": 0.00023073634132742883,
|
| 1089 |
+
"step": 15300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.3989825581395348,
|
| 1093 |
+
"grad_norm": 0.02727348543703556,
|
| 1094 |
+
"learning_rate": 1.5027707122093024e-05,
|
| 1095 |
+
"loss": 0.0003524136170744896,
|
| 1096 |
+
"step": 15400
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.4080668604651163,
|
| 1100 |
+
"grad_norm": 0.0002921252744272351,
|
| 1101 |
+
"learning_rate": 1.4800599563953488e-05,
|
| 1102 |
+
"loss": 0.000525415763258934,
|
| 1103 |
+
"step": 15500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.4171511627906976,
|
| 1107 |
+
"grad_norm": 0.0016339289722964168,
|
| 1108 |
+
"learning_rate": 1.4573492005813955e-05,
|
| 1109 |
+
"loss": 0.0003766526654362678,
|
| 1110 |
+
"step": 15600
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.426235465116279,
|
| 1114 |
+
"grad_norm": 0.003485665889456868,
|
| 1115 |
+
"learning_rate": 1.4346384447674419e-05,
|
| 1116 |
+
"loss": 0.00021764757111668586,
|
| 1117 |
+
"step": 15700
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.4353197674418605,
|
| 1121 |
+
"grad_norm": 0.0011686257785186172,
|
| 1122 |
+
"learning_rate": 1.4119276889534885e-05,
|
| 1123 |
+
"loss": 0.00026009151712059976,
|
| 1124 |
+
"step": 15800
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.4444040697674418,
|
| 1128 |
+
"grad_norm": 0.00034656753996387124,
|
| 1129 |
+
"learning_rate": 1.389216933139535e-05,
|
| 1130 |
+
"loss": 0.0003694407269358635,
|
| 1131 |
+
"step": 15900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.4534883720930232,
|
| 1135 |
+
"grad_norm": 0.0023575718514621258,
|
| 1136 |
+
"learning_rate": 1.3665061773255814e-05,
|
| 1137 |
+
"loss": 0.0001483263447880745,
|
| 1138 |
+
"step": 16000
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.4625726744186047,
|
| 1142 |
+
"grad_norm": 0.0011886970605701208,
|
| 1143 |
+
"learning_rate": 1.343795421511628e-05,
|
| 1144 |
+
"loss": 0.0004465998336672783,
|
| 1145 |
+
"step": 16100
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.471656976744186,
|
| 1149 |
+
"grad_norm": 0.00039876584196463227,
|
| 1150 |
+
"learning_rate": 1.3210846656976745e-05,
|
| 1151 |
+
"loss": 0.00015990335494279863,
|
| 1152 |
+
"step": 16200
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.4807412790697674,
|
| 1156 |
+
"grad_norm": 0.00023756893642712384,
|
| 1157 |
+
"learning_rate": 1.298373909883721e-05,
|
| 1158 |
+
"loss": 5.0160493701696394e-05,
|
| 1159 |
+
"step": 16300
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.489825581395349,
|
| 1163 |
+
"grad_norm": 0.0002809664292726666,
|
| 1164 |
+
"learning_rate": 1.2756631540697675e-05,
|
| 1165 |
+
"loss": 0.0001377291791141033,
|
| 1166 |
+
"step": 16400
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.4989098837209303,
|
| 1170 |
+
"grad_norm": 0.00037873705150559545,
|
| 1171 |
+
"learning_rate": 1.252952398255814e-05,
|
| 1172 |
+
"loss": 7.196901366114616e-05,
|
| 1173 |
+
"step": 16500
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.5079941860465116,
|
| 1177 |
+
"grad_norm": 1.4199703931808472,
|
| 1178 |
+
"learning_rate": 1.2302416424418606e-05,
|
| 1179 |
+
"loss": 0.0005905186012387276,
|
| 1180 |
+
"step": 16600
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.5170784883720931,
|
| 1184 |
+
"grad_norm": 0.0003422704176045954,
|
| 1185 |
+
"learning_rate": 1.207530886627907e-05,
|
| 1186 |
+
"loss": 0.00013428066857159137,
|
| 1187 |
+
"step": 16700
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.5261627906976745,
|
| 1191 |
+
"grad_norm": 0.0010882618371397257,
|
| 1192 |
+
"learning_rate": 1.1848201308139535e-05,
|
| 1193 |
+
"loss": 0.00020068021491169929,
|
| 1194 |
+
"step": 16800
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.5352470930232558,
|
| 1198 |
+
"grad_norm": 0.024530770257115364,
|
| 1199 |
+
"learning_rate": 1.162109375e-05,
|
| 1200 |
+
"loss": 0.0004120354726910591,
|
| 1201 |
+
"step": 16900
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.5443313953488373,
|
| 1205 |
+
"grad_norm": 0.0006738721276633441,
|
| 1206 |
+
"learning_rate": 1.1393986191860465e-05,
|
| 1207 |
+
"loss": 0.0002767092920839787,
|
| 1208 |
+
"step": 17000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.5534156976744184,
|
| 1212 |
+
"grad_norm": 0.00019564498506952077,
|
| 1213 |
+
"learning_rate": 1.1166878633720931e-05,
|
| 1214 |
+
"loss": 0.00021817052736878396,
|
| 1215 |
+
"step": 17100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.5625,
|
| 1219 |
+
"grad_norm": 0.0002832361788023263,
|
| 1220 |
+
"learning_rate": 1.0939771075581396e-05,
|
| 1221 |
+
"loss": 5.168498028069734e-05,
|
| 1222 |
+
"step": 17200
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.5715843023255816,
|
| 1226 |
+
"grad_norm": 0.0005729420809075236,
|
| 1227 |
+
"learning_rate": 1.071266351744186e-05,
|
| 1228 |
+
"loss": 0.00016681572422385215,
|
| 1229 |
+
"step": 17300
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.5806686046511627,
|
| 1233 |
+
"grad_norm": 0.0007383101619780064,
|
| 1234 |
+
"learning_rate": 1.0485555959302326e-05,
|
| 1235 |
+
"loss": 0.0003441007435321808,
|
| 1236 |
+
"step": 17400
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.5897529069767442,
|
| 1240 |
+
"grad_norm": 0.005897729191929102,
|
| 1241 |
+
"learning_rate": 1.0258448401162791e-05,
|
| 1242 |
+
"loss": 0.0003161391615867615,
|
| 1243 |
+
"step": 17500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.5988372093023255,
|
| 1247 |
+
"grad_norm": 0.00025281202397309244,
|
| 1248 |
+
"learning_rate": 1.0031340843023257e-05,
|
| 1249 |
+
"loss": 0.00011429931037127971,
|
| 1250 |
+
"step": 17600
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.6079215116279069,
|
| 1254 |
+
"grad_norm": 0.0004013874859083444,
|
| 1255 |
+
"learning_rate": 9.804233284883721e-06,
|
| 1256 |
+
"loss": 2.793062012642622e-05,
|
| 1257 |
+
"step": 17700
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.6170058139534884,
|
| 1261 |
+
"grad_norm": 0.008875502273440361,
|
| 1262 |
+
"learning_rate": 9.577125726744186e-06,
|
| 1263 |
+
"loss": 5.882895085960627e-05,
|
| 1264 |
+
"step": 17800
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.6260901162790697,
|
| 1268 |
+
"grad_norm": 0.001823420519940555,
|
| 1269 |
+
"learning_rate": 9.350018168604652e-06,
|
| 1270 |
+
"loss": 0.0002583874017000198,
|
| 1271 |
+
"step": 17900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.635174418604651,
|
| 1275 |
+
"grad_norm": 0.6227073073387146,
|
| 1276 |
+
"learning_rate": 9.122910610465116e-06,
|
| 1277 |
+
"loss": 0.0007115737348794937,
|
| 1278 |
+
"step": 18000
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.6442587209302326,
|
| 1282 |
+
"grad_norm": 0.0017345056403428316,
|
| 1283 |
+
"learning_rate": 8.895803052325581e-06,
|
| 1284 |
+
"loss": 0.00027986690402030944,
|
| 1285 |
+
"step": 18100
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.653343023255814,
|
| 1289 |
+
"grad_norm": 0.0009033643291331828,
|
| 1290 |
+
"learning_rate": 8.668695494186047e-06,
|
| 1291 |
+
"loss": 0.0002979452162981033,
|
| 1292 |
+
"step": 18200
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.6624273255813953,
|
| 1296 |
+
"grad_norm": 0.00028923453646712005,
|
| 1297 |
+
"learning_rate": 8.441587936046511e-06,
|
| 1298 |
+
"loss": 1.926603843457997e-05,
|
| 1299 |
+
"step": 18300
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.6715116279069768,
|
| 1303 |
+
"grad_norm": 0.0006103311898186803,
|
| 1304 |
+
"learning_rate": 8.214480377906978e-06,
|
| 1305 |
+
"loss": 6.331724114716053e-05,
|
| 1306 |
+
"step": 18400
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.6805959302325582,
|
| 1310 |
+
"grad_norm": 0.0002818437642417848,
|
| 1311 |
+
"learning_rate": 7.987372819767442e-06,
|
| 1312 |
+
"loss": 0.00014105773530900478,
|
| 1313 |
+
"step": 18500
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.6896802325581395,
|
| 1317 |
+
"grad_norm": 0.0011020454112440348,
|
| 1318 |
+
"learning_rate": 7.760265261627907e-06,
|
| 1319 |
+
"loss": 0.0002292773686349392,
|
| 1320 |
+
"step": 18600
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.698764534883721,
|
| 1324 |
+
"grad_norm": 0.001500884653069079,
|
| 1325 |
+
"learning_rate": 7.533157703488372e-06,
|
| 1326 |
+
"loss": 0.00011514685116708279,
|
| 1327 |
+
"step": 18700
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1.7078488372093024,
|
| 1331 |
+
"grad_norm": 0.00016000888717826456,
|
| 1332 |
+
"learning_rate": 7.306050145348838e-06,
|
| 1333 |
+
"loss": 2.0523781422525645e-05,
|
| 1334 |
+
"step": 18800
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.7169331395348837,
|
| 1338 |
+
"grad_norm": 0.003748674876987934,
|
| 1339 |
+
"learning_rate": 7.078942587209303e-06,
|
| 1340 |
+
"loss": 0.00043839264661073686,
|
| 1341 |
+
"step": 18900
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.7260174418604652,
|
| 1345 |
+
"grad_norm": 0.06772974133491516,
|
| 1346 |
+
"learning_rate": 6.8518350290697685e-06,
|
| 1347 |
+
"loss": 2.126413397490978e-05,
|
| 1348 |
+
"step": 19000
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.7351017441860463,
|
| 1352 |
+
"grad_norm": 0.0004789210797753185,
|
| 1353 |
+
"learning_rate": 6.624727470930232e-06,
|
| 1354 |
+
"loss": 0.00020929597318172455,
|
| 1355 |
+
"step": 19100
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.744186046511628,
|
| 1359 |
+
"grad_norm": 7.055519381538033e-05,
|
| 1360 |
+
"learning_rate": 6.397619912790697e-06,
|
| 1361 |
+
"loss": 0.00032072752714157104,
|
| 1362 |
+
"step": 19200
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.7532703488372094,
|
| 1366 |
+
"grad_norm": 0.00010641128028510138,
|
| 1367 |
+
"learning_rate": 6.1705123546511635e-06,
|
| 1368 |
+
"loss": 8.132393471896649e-05,
|
| 1369 |
+
"step": 19300
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.7623546511627906,
|
| 1373 |
+
"grad_norm": 0.0005615473492071033,
|
| 1374 |
+
"learning_rate": 5.943404796511629e-06,
|
| 1375 |
+
"loss": 9.162256610579788e-06,
|
| 1376 |
+
"step": 19400
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.771438953488372,
|
| 1380 |
+
"grad_norm": 0.004505404736846685,
|
| 1381 |
+
"learning_rate": 5.716297238372093e-06,
|
| 1382 |
+
"loss": 9.514865465462207e-05,
|
| 1383 |
+
"step": 19500
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.7805232558139537,
|
| 1387 |
+
"grad_norm": 0.0011890050955116749,
|
| 1388 |
+
"learning_rate": 5.4891896802325586e-06,
|
| 1389 |
+
"loss": 3.7443286273628474e-05,
|
| 1390 |
+
"step": 19600
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.7896075581395348,
|
| 1394 |
+
"grad_norm": 0.000441042153397575,
|
| 1395 |
+
"learning_rate": 5.262082122093023e-06,
|
| 1396 |
+
"loss": 5.087008234113455e-05,
|
| 1397 |
+
"step": 19700
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.7986918604651163,
|
| 1401 |
+
"grad_norm": 0.0001777316356310621,
|
| 1402 |
+
"learning_rate": 5.034974563953489e-06,
|
| 1403 |
+
"loss": 0.00017192648723721505,
|
| 1404 |
+
"step": 19800
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.8077761627906976,
|
| 1408 |
+
"grad_norm": 0.00015439293929375708,
|
| 1409 |
+
"learning_rate": 4.8078670058139536e-06,
|
| 1410 |
+
"loss": 0.0001058769691735506,
|
| 1411 |
+
"step": 19900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 1.816860465116279,
|
| 1415 |
+
"grad_norm": 0.0016349812503904104,
|
| 1416 |
+
"learning_rate": 4.580759447674419e-06,
|
| 1417 |
+
"loss": 1.0620863176882267e-05,
|
| 1418 |
+
"step": 20000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.8259447674418605,
|
| 1422 |
+
"grad_norm": 9.556530858390033e-05,
|
| 1423 |
+
"learning_rate": 4.353651889534884e-06,
|
| 1424 |
+
"loss": 0.00019107908010482789,
|
| 1425 |
+
"step": 20100
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.8350290697674418,
|
| 1429 |
+
"grad_norm": 0.0002568990457803011,
|
| 1430 |
+
"learning_rate": 4.126544331395349e-06,
|
| 1431 |
+
"loss": 0.00026563439518213273,
|
| 1432 |
+
"step": 20200
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.8441133720930232,
|
| 1436 |
+
"grad_norm": 0.5719628930091858,
|
| 1437 |
+
"learning_rate": 3.899436773255814e-06,
|
| 1438 |
+
"loss": 5.403281655162573e-05,
|
| 1439 |
+
"step": 20300
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.8531976744186047,
|
| 1443 |
+
"grad_norm": 0.017383404076099396,
|
| 1444 |
+
"learning_rate": 3.672329215116279e-06,
|
| 1445 |
+
"loss": 0.00036652404814958573,
|
| 1446 |
+
"step": 20400
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.862281976744186,
|
| 1450 |
+
"grad_norm": 0.00022910887491889298,
|
| 1451 |
+
"learning_rate": 3.4452216569767445e-06,
|
| 1452 |
+
"loss": 0.000403200164437294,
|
| 1453 |
+
"step": 20500
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.8713662790697674,
|
| 1457 |
+
"grad_norm": 0.00033295468892902136,
|
| 1458 |
+
"learning_rate": 3.2181140988372097e-06,
|
| 1459 |
+
"loss": 8.646129630506039e-05,
|
| 1460 |
+
"step": 20600
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.880450581395349,
|
| 1464 |
+
"grad_norm": 0.00012523184705059975,
|
| 1465 |
+
"learning_rate": 2.9910065406976746e-06,
|
| 1466 |
+
"loss": 0.00016864996403455733,
|
| 1467 |
+
"step": 20700
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.8895348837209303,
|
| 1471 |
+
"grad_norm": 0.00022476979938801378,
|
| 1472 |
+
"learning_rate": 2.7638989825581395e-06,
|
| 1473 |
+
"loss": 5.167209077626467e-05,
|
| 1474 |
+
"step": 20800
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.8986191860465116,
|
| 1478 |
+
"grad_norm": 0.0002389108412899077,
|
| 1479 |
+
"learning_rate": 2.5367914244186048e-06,
|
| 1480 |
+
"loss": 6.591790355741977e-05,
|
| 1481 |
+
"step": 20900
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.9077034883720931,
|
| 1485 |
+
"grad_norm": 0.0013002109481021762,
|
| 1486 |
+
"learning_rate": 2.30968386627907e-06,
|
| 1487 |
+
"loss": 0.00010993644595146179,
|
| 1488 |
+
"step": 21000
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.9167877906976745,
|
| 1492 |
+
"grad_norm": 0.13188259303569794,
|
| 1493 |
+
"learning_rate": 2.082576308139535e-06,
|
| 1494 |
+
"loss": 0.00014082306995987892,
|
| 1495 |
+
"step": 21100
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.9258720930232558,
|
| 1499 |
+
"grad_norm": 0.0010737127158790827,
|
| 1500 |
+
"learning_rate": 1.85546875e-06,
|
| 1501 |
+
"loss": 0.00030819986015558244,
|
| 1502 |
+
"step": 21200
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.9349563953488373,
|
| 1506 |
+
"grad_norm": 7.047707185847685e-05,
|
| 1507 |
+
"learning_rate": 1.628361191860465e-06,
|
| 1508 |
+
"loss": 2.0567586179822683e-05,
|
| 1509 |
+
"step": 21300
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.9440406976744184,
|
| 1513 |
+
"grad_norm": 0.0001350079692201689,
|
| 1514 |
+
"learning_rate": 1.4012536337209304e-06,
|
| 1515 |
+
"loss": 7.989832083694636e-06,
|
| 1516 |
+
"step": 21400
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.953125,
|
| 1520 |
+
"grad_norm": 0.0008243785705417395,
|
| 1521 |
+
"learning_rate": 1.1741460755813954e-06,
|
| 1522 |
+
"loss": 0.0002570002153515816,
|
| 1523 |
+
"step": 21500
|
| 1524 |
+
}
|
| 1525 |
+
],
|
| 1526 |
+
"logging_steps": 100,
|
| 1527 |
+
"max_steps": 22016,
|
| 1528 |
+
"num_input_tokens_seen": 0,
|
| 1529 |
+
"num_train_epochs": 2,
|
| 1530 |
+
"save_steps": 500,
|
| 1531 |
+
"stateful_callbacks": {
|
| 1532 |
+
"TrainerControl": {
|
| 1533 |
+
"args": {
|
| 1534 |
+
"should_epoch_stop": false,
|
| 1535 |
+
"should_evaluate": false,
|
| 1536 |
+
"should_log": false,
|
| 1537 |
+
"should_save": true,
|
| 1538 |
+
"should_training_stop": false
|
| 1539 |
+
},
|
| 1540 |
+
"attributes": {}
|
| 1541 |
+
}
|
| 1542 |
+
},
|
| 1543 |
+
"total_flos": 1.310887654981632e+16,
|
| 1544 |
+
"train_batch_size": 8,
|
| 1545 |
+
"trial_name": null,
|
| 1546 |
+
"trial_params": null
|
| 1547 |
+
}
|
models/dostoievsky_v1/checkpoint-21500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:edab2539a81b65c2a76c4c39bed219aa02358f44f3286429f83165442a1b53fa
|
| 3 |
+
size 5329
|
models/dostoievsky_v1/checkpoint-22000/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.1,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"BartForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.1,
|
| 10 |
+
"bos_token_id": 0,
|
| 11 |
+
"classif_dropout": 0.1,
|
| 12 |
+
"classifier_dropout": 0.0,
|
| 13 |
+
"d_model": 768,
|
| 14 |
+
"decoder_attention_heads": 12,
|
| 15 |
+
"decoder_ffn_dim": 3072,
|
| 16 |
+
"decoder_layerdrop": 0.0,
|
| 17 |
+
"decoder_layers": 6,
|
| 18 |
+
"decoder_start_token_id": 2,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 12,
|
| 22 |
+
"encoder_ffn_dim": 3072,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": true,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1,
|
| 39 |
+
"LABEL_2": 2
|
| 40 |
+
},
|
| 41 |
+
"max_position_embeddings": 1024,
|
| 42 |
+
"model_type": "bart",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": true,
|
| 45 |
+
"pad_token_id": 1,
|
| 46 |
+
"scale_embedding": false,
|
| 47 |
+
"task_specific_params": {
|
| 48 |
+
"summarization": {
|
| 49 |
+
"length_penalty": 1.0,
|
| 50 |
+
"max_length": 128,
|
| 51 |
+
"min_length": 12,
|
| 52 |
+
"num_beams": 4
|
| 53 |
+
},
|
| 54 |
+
"summarization_cnn": {
|
| 55 |
+
"length_penalty": 2.0,
|
| 56 |
+
"max_length": 142,
|
| 57 |
+
"min_length": 56,
|
| 58 |
+
"num_beams": 4
|
| 59 |
+
},
|
| 60 |
+
"summarization_xsum": {
|
| 61 |
+
"length_penalty": 1.0,
|
| 62 |
+
"max_length": 62,
|
| 63 |
+
"min_length": 11,
|
| 64 |
+
"num_beams": 6
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.5.0",
|
| 69 |
+
"use_cache": false,
|
| 70 |
+
"vocab_size": 50265
|
| 71 |
+
}
|
models/dostoievsky_v1/checkpoint-22000/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"early_stopping": true,
|
| 6 |
+
"eos_token_id": [
|
| 7 |
+
2
|
| 8 |
+
],
|
| 9 |
+
"forced_bos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 2,
|
| 11 |
+
"no_repeat_ngram_size": 3,
|
| 12 |
+
"num_beams": 4,
|
| 13 |
+
"pad_token_id": 1,
|
| 14 |
+
"transformers_version": "5.5.0",
|
| 15 |
+
"use_cache": true
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-22000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:633c3641654a45feb7257a935fd947b6989e946fa6299b29e37ba5e4b5ad0b0e
|
| 3 |
+
size 557912620
|
models/dostoievsky_v1/checkpoint-22000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d1e4af5742f873951b83cbc14534ab7498884479c96fcdac27b2628a59510db
|
| 3 |
+
size 1115583947
|
models/dostoievsky_v1/checkpoint-22000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f7968204ee7cf2f156da4351cf9b12e9e660827a0cbfd71d4a12e060ad798ec
|
| 3 |
+
size 14645
|
models/dostoievsky_v1/checkpoint-22000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7f093c3a5ebc5ff143ef105fa917c11d68551c7b60bb71edb210a12bc3eab0e
|
| 3 |
+
size 1383
|
models/dostoievsky_v1/checkpoint-22000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09cfa34929b7369dcf65b9d61627476e95b27fd4cd32f7aa9c5a8fa4268f140f
|
| 3 |
+
size 1465
|
models/dostoievsky_v1/checkpoint-22000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
models/dostoievsky_v1/checkpoint-22000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"mask_token": "<mask>",
|
| 10 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"sep_token": "</s>",
|
| 13 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 14 |
+
"trim_offsets": true,
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-22000/trainer_state.json
ADDED
|
@@ -0,0 +1,1582 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9985465116279069,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 22000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.009084302325581396,
|
| 14 |
+
"grad_norm": 1.9626818895339966,
|
| 15 |
+
"learning_rate": 4.9775163517441866e-05,
|
| 16 |
+
"loss": 3.2140347290039064,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.018168604651162792,
|
| 21 |
+
"grad_norm": 1.511451244354248,
|
| 22 |
+
"learning_rate": 4.9548055959302324e-05,
|
| 23 |
+
"loss": 0.01261173963546753,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.027252906976744186,
|
| 28 |
+
"grad_norm": 0.18194368481636047,
|
| 29 |
+
"learning_rate": 4.9320948401162795e-05,
|
| 30 |
+
"loss": 0.009183254837989808,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.036337209302325583,
|
| 35 |
+
"grad_norm": 0.27592894434928894,
|
| 36 |
+
"learning_rate": 4.909384084302326e-05,
|
| 37 |
+
"loss": 0.0027826914191246034,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.045421511627906974,
|
| 42 |
+
"grad_norm": 1.5584100484848022,
|
| 43 |
+
"learning_rate": 4.8866733284883724e-05,
|
| 44 |
+
"loss": 0.0029672542214393615,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05450581395348837,
|
| 49 |
+
"grad_norm": 0.02698471024632454,
|
| 50 |
+
"learning_rate": 4.863962572674419e-05,
|
| 51 |
+
"loss": 0.004296095669269561,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06359011627906977,
|
| 56 |
+
"grad_norm": 0.137993723154068,
|
| 57 |
+
"learning_rate": 4.841251816860465e-05,
|
| 58 |
+
"loss": 0.0031410756707191466,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07267441860465117,
|
| 63 |
+
"grad_norm": 0.07874622195959091,
|
| 64 |
+
"learning_rate": 4.818541061046512e-05,
|
| 65 |
+
"loss": 0.003037240505218506,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08175872093023256,
|
| 70 |
+
"grad_norm": 0.013660268858075142,
|
| 71 |
+
"learning_rate": 4.795830305232558e-05,
|
| 72 |
+
"loss": 0.0031082597374916076,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.09084302325581395,
|
| 77 |
+
"grad_norm": 0.3354911804199219,
|
| 78 |
+
"learning_rate": 4.7731195494186046e-05,
|
| 79 |
+
"loss": 0.0048285979032516475,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09992732558139535,
|
| 84 |
+
"grad_norm": 0.43667030334472656,
|
| 85 |
+
"learning_rate": 4.750408793604652e-05,
|
| 86 |
+
"loss": 0.0019270157814025878,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10901162790697674,
|
| 91 |
+
"grad_norm": 0.03849343582987785,
|
| 92 |
+
"learning_rate": 4.7276980377906975e-05,
|
| 93 |
+
"loss": 0.001011388897895813,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11809593023255814,
|
| 98 |
+
"grad_norm": 0.017985744401812553,
|
| 99 |
+
"learning_rate": 4.7049872819767446e-05,
|
| 100 |
+
"loss": 0.0018200889229774476,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12718023255813954,
|
| 105 |
+
"grad_norm": 0.05241613835096359,
|
| 106 |
+
"learning_rate": 4.682276526162791e-05,
|
| 107 |
+
"loss": 0.004675151705741882,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13626453488372092,
|
| 112 |
+
"grad_norm": 0.16119782626628876,
|
| 113 |
+
"learning_rate": 4.6595657703488375e-05,
|
| 114 |
+
"loss": 0.0025834646821022034,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.14534883720930233,
|
| 119 |
+
"grad_norm": 0.8985283374786377,
|
| 120 |
+
"learning_rate": 4.636855014534884e-05,
|
| 121 |
+
"loss": 0.001825849711894989,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.15443313953488372,
|
| 126 |
+
"grad_norm": 0.021285999566316605,
|
| 127 |
+
"learning_rate": 4.6141442587209304e-05,
|
| 128 |
+
"loss": 0.004164438843727112,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.16351744186046513,
|
| 133 |
+
"grad_norm": 0.01013926975429058,
|
| 134 |
+
"learning_rate": 4.591433502906977e-05,
|
| 135 |
+
"loss": 0.002582077383995056,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1726017441860465,
|
| 140 |
+
"grad_norm": 0.005713903810828924,
|
| 141 |
+
"learning_rate": 4.568722747093023e-05,
|
| 142 |
+
"loss": 0.0026018735766410827,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1816860465116279,
|
| 147 |
+
"grad_norm": 0.019032707437872887,
|
| 148 |
+
"learning_rate": 4.54601199127907e-05,
|
| 149 |
+
"loss": 0.0017240011692047118,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1907703488372093,
|
| 154 |
+
"grad_norm": 0.03191379830241203,
|
| 155 |
+
"learning_rate": 4.523301235465117e-05,
|
| 156 |
+
"loss": 0.0022749193012714386,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1998546511627907,
|
| 161 |
+
"grad_norm": 0.12945719063282013,
|
| 162 |
+
"learning_rate": 4.5005904796511626e-05,
|
| 163 |
+
"loss": 0.0017636509239673615,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2089389534883721,
|
| 168 |
+
"grad_norm": 0.01494428887963295,
|
| 169 |
+
"learning_rate": 4.47787972383721e-05,
|
| 170 |
+
"loss": 0.0017163331806659698,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2180232558139535,
|
| 175 |
+
"grad_norm": 0.08089974522590637,
|
| 176 |
+
"learning_rate": 4.455168968023256e-05,
|
| 177 |
+
"loss": 0.0023377402126789092,
|
| 178 |
+
"step": 2400
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.22710755813953487,
|
| 182 |
+
"grad_norm": 0.004136559087783098,
|
| 183 |
+
"learning_rate": 4.4324582122093026e-05,
|
| 184 |
+
"loss": 0.001353910267353058,
|
| 185 |
+
"step": 2500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.23619186046511628,
|
| 189 |
+
"grad_norm": 0.25111478567123413,
|
| 190 |
+
"learning_rate": 4.409747456395349e-05,
|
| 191 |
+
"loss": 0.0022939679026603697,
|
| 192 |
+
"step": 2600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.24527616279069767,
|
| 196 |
+
"grad_norm": 0.963623583316803,
|
| 197 |
+
"learning_rate": 4.3870367005813955e-05,
|
| 198 |
+
"loss": 0.0018738456070423125,
|
| 199 |
+
"step": 2700
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2543604651162791,
|
| 203 |
+
"grad_norm": 0.6530899405479431,
|
| 204 |
+
"learning_rate": 4.364325944767442e-05,
|
| 205 |
+
"loss": 0.0015843257308006287,
|
| 206 |
+
"step": 2800
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.26344476744186046,
|
| 210 |
+
"grad_norm": 0.0018010369967669249,
|
| 211 |
+
"learning_rate": 4.3416151889534884e-05,
|
| 212 |
+
"loss": 0.0006610354781150817,
|
| 213 |
+
"step": 2900
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.27252906976744184,
|
| 217 |
+
"grad_norm": 0.1644030064344406,
|
| 218 |
+
"learning_rate": 4.318904433139535e-05,
|
| 219 |
+
"loss": 0.0031197205185890197,
|
| 220 |
+
"step": 3000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.28161337209302323,
|
| 224 |
+
"grad_norm": 0.14440134167671204,
|
| 225 |
+
"learning_rate": 4.296193677325582e-05,
|
| 226 |
+
"loss": 0.0011694706976413728,
|
| 227 |
+
"step": 3100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.29069767441860467,
|
| 231 |
+
"grad_norm": 0.0124755734577775,
|
| 232 |
+
"learning_rate": 4.273482921511628e-05,
|
| 233 |
+
"loss": 0.0013250903785228729,
|
| 234 |
+
"step": 3200
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.29978197674418605,
|
| 238 |
+
"grad_norm": 0.15910762548446655,
|
| 239 |
+
"learning_rate": 4.250772165697675e-05,
|
| 240 |
+
"loss": 0.0007842753082513809,
|
| 241 |
+
"step": 3300
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.30886627906976744,
|
| 245 |
+
"grad_norm": 0.011540662497282028,
|
| 246 |
+
"learning_rate": 4.228061409883721e-05,
|
| 247 |
+
"loss": 0.0018773898482322693,
|
| 248 |
+
"step": 3400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3179505813953488,
|
| 252 |
+
"grad_norm": 0.005148735363036394,
|
| 253 |
+
"learning_rate": 4.205350654069768e-05,
|
| 254 |
+
"loss": 0.001056775525212288,
|
| 255 |
+
"step": 3500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.32703488372093026,
|
| 259 |
+
"grad_norm": 0.007442768197506666,
|
| 260 |
+
"learning_rate": 4.182639898255814e-05,
|
| 261 |
+
"loss": 0.0014348265528678894,
|
| 262 |
+
"step": 3600
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.33611918604651164,
|
| 266 |
+
"grad_norm": 0.00586783979088068,
|
| 267 |
+
"learning_rate": 4.1599291424418606e-05,
|
| 268 |
+
"loss": 0.0014780126512050629,
|
| 269 |
+
"step": 3700
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.345203488372093,
|
| 273 |
+
"grad_norm": 0.006909696385264397,
|
| 274 |
+
"learning_rate": 4.137218386627907e-05,
|
| 275 |
+
"loss": 0.0019674110412597656,
|
| 276 |
+
"step": 3800
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3542877906976744,
|
| 280 |
+
"grad_norm": 0.020171863958239555,
|
| 281 |
+
"learning_rate": 4.1145076308139535e-05,
|
| 282 |
+
"loss": 0.00249418705701828,
|
| 283 |
+
"step": 3900
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3633720930232558,
|
| 287 |
+
"grad_norm": 0.031723715364933014,
|
| 288 |
+
"learning_rate": 4.091796875e-05,
|
| 289 |
+
"loss": 0.0030515387654304503,
|
| 290 |
+
"step": 4000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.37245639534883723,
|
| 294 |
+
"grad_norm": 0.30729565024375916,
|
| 295 |
+
"learning_rate": 4.069086119186047e-05,
|
| 296 |
+
"loss": 0.0009509802609682083,
|
| 297 |
+
"step": 4100
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3815406976744186,
|
| 301 |
+
"grad_norm": 0.0021714100148528814,
|
| 302 |
+
"learning_rate": 4.046375363372093e-05,
|
| 303 |
+
"loss": 0.0013920699059963226,
|
| 304 |
+
"step": 4200
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.390625,
|
| 308 |
+
"grad_norm": 0.001017636270262301,
|
| 309 |
+
"learning_rate": 4.02366460755814e-05,
|
| 310 |
+
"loss": 0.0010310819000005722,
|
| 311 |
+
"step": 4300
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3997093023255814,
|
| 315 |
+
"grad_norm": 0.0009663284290581942,
|
| 316 |
+
"learning_rate": 4.0009538517441864e-05,
|
| 317 |
+
"loss": 0.001098722442984581,
|
| 318 |
+
"step": 4400
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.40879360465116277,
|
| 322 |
+
"grad_norm": 0.003507931251078844,
|
| 323 |
+
"learning_rate": 3.978243095930232e-05,
|
| 324 |
+
"loss": 0.0013840131461620331,
|
| 325 |
+
"step": 4500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4178779069767442,
|
| 329 |
+
"grad_norm": 0.21988199651241302,
|
| 330 |
+
"learning_rate": 3.955532340116279e-05,
|
| 331 |
+
"loss": 0.0017136169970035553,
|
| 332 |
+
"step": 4600
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4269622093023256,
|
| 336 |
+
"grad_norm": 0.0012972657568752766,
|
| 337 |
+
"learning_rate": 3.932821584302326e-05,
|
| 338 |
+
"loss": 0.0008040449023246765,
|
| 339 |
+
"step": 4700
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.436046511627907,
|
| 343 |
+
"grad_norm": 0.033857911825180054,
|
| 344 |
+
"learning_rate": 3.910110828488372e-05,
|
| 345 |
+
"loss": 0.0005662231892347336,
|
| 346 |
+
"step": 4800
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.44513081395348836,
|
| 350 |
+
"grad_norm": 0.0008128180052153766,
|
| 351 |
+
"learning_rate": 3.8874000726744187e-05,
|
| 352 |
+
"loss": 0.001040520742535591,
|
| 353 |
+
"step": 4900
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.45421511627906974,
|
| 357 |
+
"grad_norm": 0.0017111338675022125,
|
| 358 |
+
"learning_rate": 3.864689316860465e-05,
|
| 359 |
+
"loss": 0.0010922805964946748,
|
| 360 |
+
"step": 5000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.4632994186046512,
|
| 364 |
+
"grad_norm": 0.0013291583163663745,
|
| 365 |
+
"learning_rate": 3.841978561046512e-05,
|
| 366 |
+
"loss": 0.0005277743935585021,
|
| 367 |
+
"step": 5100
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.47238372093023256,
|
| 371 |
+
"grad_norm": 0.09846807271242142,
|
| 372 |
+
"learning_rate": 3.819267805232558e-05,
|
| 373 |
+
"loss": 0.0010997675359249114,
|
| 374 |
+
"step": 5200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.48146802325581395,
|
| 378 |
+
"grad_norm": 0.005185174290090799,
|
| 379 |
+
"learning_rate": 3.796557049418605e-05,
|
| 380 |
+
"loss": 0.000698111355304718,
|
| 381 |
+
"step": 5300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.49055232558139533,
|
| 385 |
+
"grad_norm": 0.019086388871073723,
|
| 386 |
+
"learning_rate": 3.7738462936046515e-05,
|
| 387 |
+
"loss": 0.0009016367793083191,
|
| 388 |
+
"step": 5400
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.49963662790697677,
|
| 392 |
+
"grad_norm": 0.05122831463813782,
|
| 393 |
+
"learning_rate": 3.751135537790697e-05,
|
| 394 |
+
"loss": 0.0015524370968341826,
|
| 395 |
+
"step": 5500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5087209302325582,
|
| 399 |
+
"grad_norm": 0.012596211396157742,
|
| 400 |
+
"learning_rate": 3.7284247819767444e-05,
|
| 401 |
+
"loss": 0.00220917209982872,
|
| 402 |
+
"step": 5600
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5178052325581395,
|
| 406 |
+
"grad_norm": 0.009361029602587223,
|
| 407 |
+
"learning_rate": 3.705714026162791e-05,
|
| 408 |
+
"loss": 0.0010362663865089417,
|
| 409 |
+
"step": 5700
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5268895348837209,
|
| 413 |
+
"grad_norm": 0.5205378532409668,
|
| 414 |
+
"learning_rate": 3.683003270348837e-05,
|
| 415 |
+
"loss": 0.0022550773620605468,
|
| 416 |
+
"step": 5800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5359738372093024,
|
| 420 |
+
"grad_norm": 0.0006470124353654683,
|
| 421 |
+
"learning_rate": 3.660292514534884e-05,
|
| 422 |
+
"loss": 0.0011264414340257645,
|
| 423 |
+
"step": 5900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5450581395348837,
|
| 427 |
+
"grad_norm": 0.058231666684150696,
|
| 428 |
+
"learning_rate": 3.63758175872093e-05,
|
| 429 |
+
"loss": 0.0011266635358333588,
|
| 430 |
+
"step": 6000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5541424418604651,
|
| 434 |
+
"grad_norm": 0.0013441125629469752,
|
| 435 |
+
"learning_rate": 3.614871002906977e-05,
|
| 436 |
+
"loss": 0.0008001205325126648,
|
| 437 |
+
"step": 6100
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5632267441860465,
|
| 441 |
+
"grad_norm": 0.0021539030130952597,
|
| 442 |
+
"learning_rate": 3.592160247093023e-05,
|
| 443 |
+
"loss": 0.0008849448710680007,
|
| 444 |
+
"step": 6200
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5723110465116279,
|
| 448 |
+
"grad_norm": 0.9019960761070251,
|
| 449 |
+
"learning_rate": 3.56944949127907e-05,
|
| 450 |
+
"loss": 0.0006517694145441055,
|
| 451 |
+
"step": 6300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5813953488372093,
|
| 455 |
+
"grad_norm": 0.001283760997466743,
|
| 456 |
+
"learning_rate": 3.546738735465117e-05,
|
| 457 |
+
"loss": 0.0021727623045444486,
|
| 458 |
+
"step": 6400
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5904796511627907,
|
| 462 |
+
"grad_norm": 0.005666423588991165,
|
| 463 |
+
"learning_rate": 3.5240279796511624e-05,
|
| 464 |
+
"loss": 0.0012587083876132966,
|
| 465 |
+
"step": 6500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5995639534883721,
|
| 469 |
+
"grad_norm": 1.4842944145202637,
|
| 470 |
+
"learning_rate": 3.5013172238372096e-05,
|
| 471 |
+
"loss": 0.001180112287402153,
|
| 472 |
+
"step": 6600
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6086482558139535,
|
| 476 |
+
"grad_norm": 0.02973734401166439,
|
| 477 |
+
"learning_rate": 3.478606468023256e-05,
|
| 478 |
+
"loss": 0.0010163599252700805,
|
| 479 |
+
"step": 6700
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6177325581395349,
|
| 483 |
+
"grad_norm": 0.00399527233093977,
|
| 484 |
+
"learning_rate": 3.4558957122093024e-05,
|
| 485 |
+
"loss": 0.0010414445400238038,
|
| 486 |
+
"step": 6800
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6268168604651163,
|
| 490 |
+
"grad_norm": 0.39672717452049255,
|
| 491 |
+
"learning_rate": 3.433184956395349e-05,
|
| 492 |
+
"loss": 0.0012094499170780182,
|
| 493 |
+
"step": 6900
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6359011627906976,
|
| 497 |
+
"grad_norm": 0.5308993458747864,
|
| 498 |
+
"learning_rate": 3.410474200581395e-05,
|
| 499 |
+
"loss": 0.0006021633744239807,
|
| 500 |
+
"step": 7000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6449854651162791,
|
| 504 |
+
"grad_norm": 0.003343795659020543,
|
| 505 |
+
"learning_rate": 3.3877634447674425e-05,
|
| 506 |
+
"loss": 0.0005116893351078033,
|
| 507 |
+
"step": 7100
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6540697674418605,
|
| 511 |
+
"grad_norm": 0.0033038391266018152,
|
| 512 |
+
"learning_rate": 3.365052688953488e-05,
|
| 513 |
+
"loss": 0.002106922417879105,
|
| 514 |
+
"step": 7200
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6631540697674418,
|
| 518 |
+
"grad_norm": 0.004004980903118849,
|
| 519 |
+
"learning_rate": 3.3423419331395353e-05,
|
| 520 |
+
"loss": 0.0010233993828296661,
|
| 521 |
+
"step": 7300
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6722383720930233,
|
| 525 |
+
"grad_norm": 0.00490298168733716,
|
| 526 |
+
"learning_rate": 3.319631177325582e-05,
|
| 527 |
+
"loss": 0.0007062336057424546,
|
| 528 |
+
"step": 7400
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6813226744186046,
|
| 532 |
+
"grad_norm": 0.04243200644850731,
|
| 533 |
+
"learning_rate": 3.2969204215116276e-05,
|
| 534 |
+
"loss": 0.0005479569733142852,
|
| 535 |
+
"step": 7500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.690406976744186,
|
| 539 |
+
"grad_norm": 0.008556894026696682,
|
| 540 |
+
"learning_rate": 3.274209665697675e-05,
|
| 541 |
+
"loss": 0.0010938134789466858,
|
| 542 |
+
"step": 7600
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6994912790697675,
|
| 546 |
+
"grad_norm": 0.016938723623752594,
|
| 547 |
+
"learning_rate": 3.251498909883721e-05,
|
| 548 |
+
"loss": 0.0009466408193111419,
|
| 549 |
+
"step": 7700
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7085755813953488,
|
| 553 |
+
"grad_norm": 0.09418574720621109,
|
| 554 |
+
"learning_rate": 3.2287881540697676e-05,
|
| 555 |
+
"loss": 0.00145876482129097,
|
| 556 |
+
"step": 7800
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7176598837209303,
|
| 560 |
+
"grad_norm": 0.002359782112762332,
|
| 561 |
+
"learning_rate": 3.206077398255814e-05,
|
| 562 |
+
"loss": 0.0015742655098438263,
|
| 563 |
+
"step": 7900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7267441860465116,
|
| 567 |
+
"grad_norm": 0.3374776840209961,
|
| 568 |
+
"learning_rate": 3.1833666424418605e-05,
|
| 569 |
+
"loss": 0.0007466593384742737,
|
| 570 |
+
"step": 8000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.735828488372093,
|
| 574 |
+
"grad_norm": 0.000735403154976666,
|
| 575 |
+
"learning_rate": 3.1606558866279076e-05,
|
| 576 |
+
"loss": 0.0008171546459197998,
|
| 577 |
+
"step": 8100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7449127906976745,
|
| 581 |
+
"grad_norm": 0.01963644102215767,
|
| 582 |
+
"learning_rate": 3.1379451308139533e-05,
|
| 583 |
+
"loss": 0.0007363802939653396,
|
| 584 |
+
"step": 8200
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7539970930232558,
|
| 588 |
+
"grad_norm": 0.09964141249656677,
|
| 589 |
+
"learning_rate": 3.1152343750000005e-05,
|
| 590 |
+
"loss": 0.00030826406553387644,
|
| 591 |
+
"step": 8300
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7630813953488372,
|
| 595 |
+
"grad_norm": 0.0029934593476355076,
|
| 596 |
+
"learning_rate": 3.092523619186047e-05,
|
| 597 |
+
"loss": 0.0003284827247262001,
|
| 598 |
+
"step": 8400
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7721656976744186,
|
| 602 |
+
"grad_norm": 0.002162993187084794,
|
| 603 |
+
"learning_rate": 3.069812863372093e-05,
|
| 604 |
+
"loss": 0.0010288888961076737,
|
| 605 |
+
"step": 8500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.78125,
|
| 609 |
+
"grad_norm": 0.013634397648274899,
|
| 610 |
+
"learning_rate": 3.0471021075581398e-05,
|
| 611 |
+
"loss": 0.0006983913481235504,
|
| 612 |
+
"step": 8600
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7903343023255814,
|
| 616 |
+
"grad_norm": 0.0833001658320427,
|
| 617 |
+
"learning_rate": 3.0243913517441862e-05,
|
| 618 |
+
"loss": 0.0009967025369405746,
|
| 619 |
+
"step": 8700
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7994186046511628,
|
| 623 |
+
"grad_norm": 0.4697379767894745,
|
| 624 |
+
"learning_rate": 3.0016805959302323e-05,
|
| 625 |
+
"loss": 0.0007716407626867294,
|
| 626 |
+
"step": 8800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8085029069767442,
|
| 630 |
+
"grad_norm": 0.04588587209582329,
|
| 631 |
+
"learning_rate": 2.978969840116279e-05,
|
| 632 |
+
"loss": 0.000263803955167532,
|
| 633 |
+
"step": 8900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8175872093023255,
|
| 637 |
+
"grad_norm": 0.003420337103307247,
|
| 638 |
+
"learning_rate": 2.956259084302326e-05,
|
| 639 |
+
"loss": 0.0009599439054727555,
|
| 640 |
+
"step": 9000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.826671511627907,
|
| 644 |
+
"grad_norm": 0.0024436134845018387,
|
| 645 |
+
"learning_rate": 2.9335483284883724e-05,
|
| 646 |
+
"loss": 0.012270723581314086,
|
| 647 |
+
"step": 9100
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8357558139534884,
|
| 651 |
+
"grad_norm": 0.001418731757439673,
|
| 652 |
+
"learning_rate": 2.9108375726744185e-05,
|
| 653 |
+
"loss": 0.0009713788330554962,
|
| 654 |
+
"step": 9200
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8448401162790697,
|
| 658 |
+
"grad_norm": 0.0026375153101980686,
|
| 659 |
+
"learning_rate": 2.8881268168604652e-05,
|
| 660 |
+
"loss": 0.0007654589414596558,
|
| 661 |
+
"step": 9300
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8539244186046512,
|
| 665 |
+
"grad_norm": 0.0013760777655988932,
|
| 666 |
+
"learning_rate": 2.865416061046512e-05,
|
| 667 |
+
"loss": 0.0006276721507310867,
|
| 668 |
+
"step": 9400
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8630087209302325,
|
| 672 |
+
"grad_norm": 0.0033582421019673347,
|
| 673 |
+
"learning_rate": 2.842705305232558e-05,
|
| 674 |
+
"loss": 0.0003520375117659569,
|
| 675 |
+
"step": 9500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.872093023255814,
|
| 679 |
+
"grad_norm": 0.0043012769892811775,
|
| 680 |
+
"learning_rate": 2.819994549418605e-05,
|
| 681 |
+
"loss": 0.0011182524263858794,
|
| 682 |
+
"step": 9600
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8811773255813954,
|
| 686 |
+
"grad_norm": 0.0015025343745946884,
|
| 687 |
+
"learning_rate": 2.7972837936046514e-05,
|
| 688 |
+
"loss": 0.00044553544372320173,
|
| 689 |
+
"step": 9700
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8902616279069767,
|
| 693 |
+
"grad_norm": 0.0018241156358271837,
|
| 694 |
+
"learning_rate": 2.7745730377906975e-05,
|
| 695 |
+
"loss": 0.00023134740069508553,
|
| 696 |
+
"step": 9800
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8993459302325582,
|
| 700 |
+
"grad_norm": 0.001082456554286182,
|
| 701 |
+
"learning_rate": 2.7518622819767442e-05,
|
| 702 |
+
"loss": 0.001018296480178833,
|
| 703 |
+
"step": 9900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.9084302325581395,
|
| 707 |
+
"grad_norm": 0.01155087724328041,
|
| 708 |
+
"learning_rate": 2.729151526162791e-05,
|
| 709 |
+
"loss": 0.0005589094385504722,
|
| 710 |
+
"step": 10000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.9175145348837209,
|
| 714 |
+
"grad_norm": 0.0022482366766780615,
|
| 715 |
+
"learning_rate": 2.7064407703488375e-05,
|
| 716 |
+
"loss": 0.000512191392481327,
|
| 717 |
+
"step": 10100
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9265988372093024,
|
| 721 |
+
"grad_norm": 0.20549768209457397,
|
| 722 |
+
"learning_rate": 2.6837300145348836e-05,
|
| 723 |
+
"loss": 0.00038118865340948106,
|
| 724 |
+
"step": 10200
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9356831395348837,
|
| 728 |
+
"grad_norm": 0.0013188497396185994,
|
| 729 |
+
"learning_rate": 2.6610192587209304e-05,
|
| 730 |
+
"loss": 0.0003574254736304283,
|
| 731 |
+
"step": 10300
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9447674418604651,
|
| 735 |
+
"grad_norm": 0.0006103936466388404,
|
| 736 |
+
"learning_rate": 2.638308502906977e-05,
|
| 737 |
+
"loss": 0.00024356411769986154,
|
| 738 |
+
"step": 10400
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9538517441860465,
|
| 742 |
+
"grad_norm": 0.00047453015577048063,
|
| 743 |
+
"learning_rate": 2.6155977470930233e-05,
|
| 744 |
+
"loss": 0.0013259868323802948,
|
| 745 |
+
"step": 10500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9629360465116279,
|
| 749 |
+
"grad_norm": 0.025760261341929436,
|
| 750 |
+
"learning_rate": 2.59288699127907e-05,
|
| 751 |
+
"loss": 0.0006390263140201569,
|
| 752 |
+
"step": 10600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9720203488372093,
|
| 756 |
+
"grad_norm": 0.0025282115675508976,
|
| 757 |
+
"learning_rate": 2.5701762354651165e-05,
|
| 758 |
+
"loss": 0.001759422868490219,
|
| 759 |
+
"step": 10700
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.9811046511627907,
|
| 763 |
+
"grad_norm": 2.990682601928711,
|
| 764 |
+
"learning_rate": 2.5474654796511626e-05,
|
| 765 |
+
"loss": 0.00041438560932874677,
|
| 766 |
+
"step": 10800
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9901889534883721,
|
| 770 |
+
"grad_norm": 0.0036354295443743467,
|
| 771 |
+
"learning_rate": 2.5247547238372094e-05,
|
| 772 |
+
"loss": 0.000467718243598938,
|
| 773 |
+
"step": 10900
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9992732558139535,
|
| 777 |
+
"grad_norm": 0.02648981101810932,
|
| 778 |
+
"learning_rate": 2.502043968023256e-05,
|
| 779 |
+
"loss": 0.0007214382290840149,
|
| 780 |
+
"step": 11000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0,
|
| 784 |
+
"eval_loss": 8.017317304620519e-05,
|
| 785 |
+
"eval_runtime": 132.6261,
|
| 786 |
+
"eval_samples_per_second": 165.993,
|
| 787 |
+
"eval_steps_per_second": 20.75,
|
| 788 |
+
"step": 11008
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.008357558139535,
|
| 792 |
+
"grad_norm": 0.0011333210859447718,
|
| 793 |
+
"learning_rate": 2.4793332122093026e-05,
|
| 794 |
+
"loss": 0.000609114095568657,
|
| 795 |
+
"step": 11100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.0174418604651163,
|
| 799 |
+
"grad_norm": 0.0006669509457424283,
|
| 800 |
+
"learning_rate": 2.4566224563953487e-05,
|
| 801 |
+
"loss": 0.000269341841340065,
|
| 802 |
+
"step": 11200
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.0265261627906976,
|
| 806 |
+
"grad_norm": 0.0006085751811042428,
|
| 807 |
+
"learning_rate": 2.4339117005813955e-05,
|
| 808 |
+
"loss": 0.00018008088693022728,
|
| 809 |
+
"step": 11300
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.035610465116279,
|
| 813 |
+
"grad_norm": 0.00029874974279664457,
|
| 814 |
+
"learning_rate": 2.411200944767442e-05,
|
| 815 |
+
"loss": 8.143479935824871e-05,
|
| 816 |
+
"step": 11400
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.0446947674418605,
|
| 820 |
+
"grad_norm": 0.0020065035205334425,
|
| 821 |
+
"learning_rate": 2.3884901889534887e-05,
|
| 822 |
+
"loss": 0.0004405264183878899,
|
| 823 |
+
"step": 11500
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.0537790697674418,
|
| 827 |
+
"grad_norm": 0.3139957785606384,
|
| 828 |
+
"learning_rate": 2.3657794331395348e-05,
|
| 829 |
+
"loss": 0.0007688279449939728,
|
| 830 |
+
"step": 11600
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.0628633720930232,
|
| 834 |
+
"grad_norm": 0.014869201928377151,
|
| 835 |
+
"learning_rate": 2.3430686773255813e-05,
|
| 836 |
+
"loss": 0.00045145414769649507,
|
| 837 |
+
"step": 11700
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.0719476744186047,
|
| 841 |
+
"grad_norm": 0.03592238947749138,
|
| 842 |
+
"learning_rate": 2.320357921511628e-05,
|
| 843 |
+
"loss": 0.0013660797476768493,
|
| 844 |
+
"step": 11800
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.081031976744186,
|
| 848 |
+
"grad_norm": 0.010270297527313232,
|
| 849 |
+
"learning_rate": 2.2976471656976745e-05,
|
| 850 |
+
"loss": 0.0005171676725149155,
|
| 851 |
+
"step": 11900
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.0901162790697674,
|
| 855 |
+
"grad_norm": 0.0026885548140853643,
|
| 856 |
+
"learning_rate": 2.2749364098837213e-05,
|
| 857 |
+
"loss": 0.00046057451516389846,
|
| 858 |
+
"step": 12000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.099200581395349,
|
| 862 |
+
"grad_norm": 0.005681134294718504,
|
| 863 |
+
"learning_rate": 2.2522256540697674e-05,
|
| 864 |
+
"loss": 0.0006235280632972718,
|
| 865 |
+
"step": 12100
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.1082848837209303,
|
| 869 |
+
"grad_norm": 0.0028001824393868446,
|
| 870 |
+
"learning_rate": 2.2295148982558138e-05,
|
| 871 |
+
"loss": 0.0008198145776987075,
|
| 872 |
+
"step": 12200
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.1173691860465116,
|
| 876 |
+
"grad_norm": 0.0006385694723576307,
|
| 877 |
+
"learning_rate": 2.2068041424418606e-05,
|
| 878 |
+
"loss": 0.0003722207620739937,
|
| 879 |
+
"step": 12300
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.1264534883720931,
|
| 883 |
+
"grad_norm": 0.0008649047813378274,
|
| 884 |
+
"learning_rate": 2.184093386627907e-05,
|
| 885 |
+
"loss": 0.0003394853696227074,
|
| 886 |
+
"step": 12400
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.1355377906976745,
|
| 890 |
+
"grad_norm": 0.004796006251126528,
|
| 891 |
+
"learning_rate": 2.1613826308139538e-05,
|
| 892 |
+
"loss": 0.00042502928525209424,
|
| 893 |
+
"step": 12500
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.1446220930232558,
|
| 897 |
+
"grad_norm": 0.001498043886385858,
|
| 898 |
+
"learning_rate": 2.138671875e-05,
|
| 899 |
+
"loss": 0.0009064202010631562,
|
| 900 |
+
"step": 12600
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.1537063953488371,
|
| 904 |
+
"grad_norm": 0.01664622500538826,
|
| 905 |
+
"learning_rate": 2.1159611191860464e-05,
|
| 906 |
+
"loss": 0.0002878718450665474,
|
| 907 |
+
"step": 12700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.1627906976744187,
|
| 911 |
+
"grad_norm": 0.010639426298439503,
|
| 912 |
+
"learning_rate": 2.093250363372093e-05,
|
| 913 |
+
"loss": 0.0011860337108373642,
|
| 914 |
+
"step": 12800
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.171875,
|
| 918 |
+
"grad_norm": 0.050644513219594955,
|
| 919 |
+
"learning_rate": 2.0705396075581396e-05,
|
| 920 |
+
"loss": 0.0006260576844215393,
|
| 921 |
+
"step": 12900
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.1809593023255813,
|
| 925 |
+
"grad_norm": 0.001398528809659183,
|
| 926 |
+
"learning_rate": 2.0478288517441864e-05,
|
| 927 |
+
"loss": 0.0001833665184676647,
|
| 928 |
+
"step": 13000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.1900436046511629,
|
| 932 |
+
"grad_norm": 0.0057320562191307545,
|
| 933 |
+
"learning_rate": 2.0251180959302325e-05,
|
| 934 |
+
"loss": 0.00032943256199359896,
|
| 935 |
+
"step": 13100
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.1991279069767442,
|
| 939 |
+
"grad_norm": 0.0002997924748342484,
|
| 940 |
+
"learning_rate": 2.002407340116279e-05,
|
| 941 |
+
"loss": 6.249572150409222e-05,
|
| 942 |
+
"step": 13200
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.2082122093023255,
|
| 946 |
+
"grad_norm": 0.0008130258647724986,
|
| 947 |
+
"learning_rate": 1.9796965843023257e-05,
|
| 948 |
+
"loss": 0.00026577839627861977,
|
| 949 |
+
"step": 13300
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.2172965116279069,
|
| 953 |
+
"grad_norm": 0.0006040658336132765,
|
| 954 |
+
"learning_rate": 1.956985828488372e-05,
|
| 955 |
+
"loss": 0.00023248378187417984,
|
| 956 |
+
"step": 13400
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.2263808139534884,
|
| 960 |
+
"grad_norm": 0.000803643895778805,
|
| 961 |
+
"learning_rate": 1.934275072674419e-05,
|
| 962 |
+
"loss": 0.0002730824239552021,
|
| 963 |
+
"step": 13500
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.2354651162790697,
|
| 967 |
+
"grad_norm": 0.0005482266424223781,
|
| 968 |
+
"learning_rate": 1.911564316860465e-05,
|
| 969 |
+
"loss": 0.0005848826467990875,
|
| 970 |
+
"step": 13600
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.244549418604651,
|
| 974 |
+
"grad_norm": 0.0034759771078824997,
|
| 975 |
+
"learning_rate": 1.8888535610465115e-05,
|
| 976 |
+
"loss": 0.0005462893471121788,
|
| 977 |
+
"step": 13700
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.2536337209302326,
|
| 981 |
+
"grad_norm": 0.0002226918877568096,
|
| 982 |
+
"learning_rate": 1.8661428052325583e-05,
|
| 983 |
+
"loss": 7.945694960653782e-05,
|
| 984 |
+
"step": 13800
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.262718023255814,
|
| 988 |
+
"grad_norm": 0.002374310279265046,
|
| 989 |
+
"learning_rate": 1.8434320494186047e-05,
|
| 990 |
+
"loss": 0.000744214728474617,
|
| 991 |
+
"step": 13900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.2718023255813953,
|
| 995 |
+
"grad_norm": 0.0009076696587726474,
|
| 996 |
+
"learning_rate": 1.8207212936046515e-05,
|
| 997 |
+
"loss": 0.0003250580281019211,
|
| 998 |
+
"step": 14000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.2808866279069768,
|
| 1002 |
+
"grad_norm": 0.010181100107729435,
|
| 1003 |
+
"learning_rate": 1.7980105377906976e-05,
|
| 1004 |
+
"loss": 0.0005765938758850097,
|
| 1005 |
+
"step": 14100
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.2899709302325582,
|
| 1009 |
+
"grad_norm": 0.0012442917795851827,
|
| 1010 |
+
"learning_rate": 1.775299781976744e-05,
|
| 1011 |
+
"loss": 0.00032320961356163024,
|
| 1012 |
+
"step": 14200
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.2990552325581395,
|
| 1016 |
+
"grad_norm": 0.0005617383285425603,
|
| 1017 |
+
"learning_rate": 1.752589026162791e-05,
|
| 1018 |
+
"loss": 0.0001776321791112423,
|
| 1019 |
+
"step": 14300
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.308139534883721,
|
| 1023 |
+
"grad_norm": 0.0010064981179311872,
|
| 1024 |
+
"learning_rate": 1.7298782703488373e-05,
|
| 1025 |
+
"loss": 0.00012774170376360415,
|
| 1026 |
+
"step": 14400
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.3172238372093024,
|
| 1030 |
+
"grad_norm": 0.00039220438338816166,
|
| 1031 |
+
"learning_rate": 1.707167514534884e-05,
|
| 1032 |
+
"loss": 0.0006190959364175796,
|
| 1033 |
+
"step": 14500
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.3263081395348837,
|
| 1037 |
+
"grad_norm": 0.0009137202869169414,
|
| 1038 |
+
"learning_rate": 1.6844567587209302e-05,
|
| 1039 |
+
"loss": 0.0007051125913858414,
|
| 1040 |
+
"step": 14600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.3353924418604652,
|
| 1044 |
+
"grad_norm": 0.0006007241318002343,
|
| 1045 |
+
"learning_rate": 1.6617460029069766e-05,
|
| 1046 |
+
"loss": 7.074063178151846e-05,
|
| 1047 |
+
"step": 14700
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.3444767441860466,
|
| 1051 |
+
"grad_norm": 0.0006039150175638497,
|
| 1052 |
+
"learning_rate": 1.6390352470930234e-05,
|
| 1053 |
+
"loss": 0.000489293597638607,
|
| 1054 |
+
"step": 14800
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.353561046511628,
|
| 1058 |
+
"grad_norm": 0.0014955669175833464,
|
| 1059 |
+
"learning_rate": 1.61632449127907e-05,
|
| 1060 |
+
"loss": 0.0004155828058719635,
|
| 1061 |
+
"step": 14900
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.3626453488372092,
|
| 1065 |
+
"grad_norm": 0.1508745402097702,
|
| 1066 |
+
"learning_rate": 1.5936137354651163e-05,
|
| 1067 |
+
"loss": 0.00024092141538858413,
|
| 1068 |
+
"step": 15000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.3717296511627908,
|
| 1072 |
+
"grad_norm": 0.0008113393560051918,
|
| 1073 |
+
"learning_rate": 1.5709029796511627e-05,
|
| 1074 |
+
"loss": 0.00040562458336353304,
|
| 1075 |
+
"step": 15100
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.380813953488372,
|
| 1079 |
+
"grad_norm": 0.0026459796354174614,
|
| 1080 |
+
"learning_rate": 1.5481922238372092e-05,
|
| 1081 |
+
"loss": 0.0007934534549713135,
|
| 1082 |
+
"step": 15200
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.3898982558139534,
|
| 1086 |
+
"grad_norm": 0.0005005365237593651,
|
| 1087 |
+
"learning_rate": 1.525481468023256e-05,
|
| 1088 |
+
"loss": 0.00023073634132742883,
|
| 1089 |
+
"step": 15300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.3989825581395348,
|
| 1093 |
+
"grad_norm": 0.02727348543703556,
|
| 1094 |
+
"learning_rate": 1.5027707122093024e-05,
|
| 1095 |
+
"loss": 0.0003524136170744896,
|
| 1096 |
+
"step": 15400
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.4080668604651163,
|
| 1100 |
+
"grad_norm": 0.0002921252744272351,
|
| 1101 |
+
"learning_rate": 1.4800599563953488e-05,
|
| 1102 |
+
"loss": 0.000525415763258934,
|
| 1103 |
+
"step": 15500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.4171511627906976,
|
| 1107 |
+
"grad_norm": 0.0016339289722964168,
|
| 1108 |
+
"learning_rate": 1.4573492005813955e-05,
|
| 1109 |
+
"loss": 0.0003766526654362678,
|
| 1110 |
+
"step": 15600
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.426235465116279,
|
| 1114 |
+
"grad_norm": 0.003485665889456868,
|
| 1115 |
+
"learning_rate": 1.4346384447674419e-05,
|
| 1116 |
+
"loss": 0.00021764757111668586,
|
| 1117 |
+
"step": 15700
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.4353197674418605,
|
| 1121 |
+
"grad_norm": 0.0011686257785186172,
|
| 1122 |
+
"learning_rate": 1.4119276889534885e-05,
|
| 1123 |
+
"loss": 0.00026009151712059976,
|
| 1124 |
+
"step": 15800
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.4444040697674418,
|
| 1128 |
+
"grad_norm": 0.00034656753996387124,
|
| 1129 |
+
"learning_rate": 1.389216933139535e-05,
|
| 1130 |
+
"loss": 0.0003694407269358635,
|
| 1131 |
+
"step": 15900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.4534883720930232,
|
| 1135 |
+
"grad_norm": 0.0023575718514621258,
|
| 1136 |
+
"learning_rate": 1.3665061773255814e-05,
|
| 1137 |
+
"loss": 0.0001483263447880745,
|
| 1138 |
+
"step": 16000
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.4625726744186047,
|
| 1142 |
+
"grad_norm": 0.0011886970605701208,
|
| 1143 |
+
"learning_rate": 1.343795421511628e-05,
|
| 1144 |
+
"loss": 0.0004465998336672783,
|
| 1145 |
+
"step": 16100
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.471656976744186,
|
| 1149 |
+
"grad_norm": 0.00039876584196463227,
|
| 1150 |
+
"learning_rate": 1.3210846656976745e-05,
|
| 1151 |
+
"loss": 0.00015990335494279863,
|
| 1152 |
+
"step": 16200
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.4807412790697674,
|
| 1156 |
+
"grad_norm": 0.00023756893642712384,
|
| 1157 |
+
"learning_rate": 1.298373909883721e-05,
|
| 1158 |
+
"loss": 5.0160493701696394e-05,
|
| 1159 |
+
"step": 16300
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.489825581395349,
|
| 1163 |
+
"grad_norm": 0.0002809664292726666,
|
| 1164 |
+
"learning_rate": 1.2756631540697675e-05,
|
| 1165 |
+
"loss": 0.0001377291791141033,
|
| 1166 |
+
"step": 16400
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.4989098837209303,
|
| 1170 |
+
"grad_norm": 0.00037873705150559545,
|
| 1171 |
+
"learning_rate": 1.252952398255814e-05,
|
| 1172 |
+
"loss": 7.196901366114616e-05,
|
| 1173 |
+
"step": 16500
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.5079941860465116,
|
| 1177 |
+
"grad_norm": 1.4199703931808472,
|
| 1178 |
+
"learning_rate": 1.2302416424418606e-05,
|
| 1179 |
+
"loss": 0.0005905186012387276,
|
| 1180 |
+
"step": 16600
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.5170784883720931,
|
| 1184 |
+
"grad_norm": 0.0003422704176045954,
|
| 1185 |
+
"learning_rate": 1.207530886627907e-05,
|
| 1186 |
+
"loss": 0.00013428066857159137,
|
| 1187 |
+
"step": 16700
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.5261627906976745,
|
| 1191 |
+
"grad_norm": 0.0010882618371397257,
|
| 1192 |
+
"learning_rate": 1.1848201308139535e-05,
|
| 1193 |
+
"loss": 0.00020068021491169929,
|
| 1194 |
+
"step": 16800
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.5352470930232558,
|
| 1198 |
+
"grad_norm": 0.024530770257115364,
|
| 1199 |
+
"learning_rate": 1.162109375e-05,
|
| 1200 |
+
"loss": 0.0004120354726910591,
|
| 1201 |
+
"step": 16900
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.5443313953488373,
|
| 1205 |
+
"grad_norm": 0.0006738721276633441,
|
| 1206 |
+
"learning_rate": 1.1393986191860465e-05,
|
| 1207 |
+
"loss": 0.0002767092920839787,
|
| 1208 |
+
"step": 17000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.5534156976744184,
|
| 1212 |
+
"grad_norm": 0.00019564498506952077,
|
| 1213 |
+
"learning_rate": 1.1166878633720931e-05,
|
| 1214 |
+
"loss": 0.00021817052736878396,
|
| 1215 |
+
"step": 17100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.5625,
|
| 1219 |
+
"grad_norm": 0.0002832361788023263,
|
| 1220 |
+
"learning_rate": 1.0939771075581396e-05,
|
| 1221 |
+
"loss": 5.168498028069734e-05,
|
| 1222 |
+
"step": 17200
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.5715843023255816,
|
| 1226 |
+
"grad_norm": 0.0005729420809075236,
|
| 1227 |
+
"learning_rate": 1.071266351744186e-05,
|
| 1228 |
+
"loss": 0.00016681572422385215,
|
| 1229 |
+
"step": 17300
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.5806686046511627,
|
| 1233 |
+
"grad_norm": 0.0007383101619780064,
|
| 1234 |
+
"learning_rate": 1.0485555959302326e-05,
|
| 1235 |
+
"loss": 0.0003441007435321808,
|
| 1236 |
+
"step": 17400
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.5897529069767442,
|
| 1240 |
+
"grad_norm": 0.005897729191929102,
|
| 1241 |
+
"learning_rate": 1.0258448401162791e-05,
|
| 1242 |
+
"loss": 0.0003161391615867615,
|
| 1243 |
+
"step": 17500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.5988372093023255,
|
| 1247 |
+
"grad_norm": 0.00025281202397309244,
|
| 1248 |
+
"learning_rate": 1.0031340843023257e-05,
|
| 1249 |
+
"loss": 0.00011429931037127971,
|
| 1250 |
+
"step": 17600
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.6079215116279069,
|
| 1254 |
+
"grad_norm": 0.0004013874859083444,
|
| 1255 |
+
"learning_rate": 9.804233284883721e-06,
|
| 1256 |
+
"loss": 2.793062012642622e-05,
|
| 1257 |
+
"step": 17700
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.6170058139534884,
|
| 1261 |
+
"grad_norm": 0.008875502273440361,
|
| 1262 |
+
"learning_rate": 9.577125726744186e-06,
|
| 1263 |
+
"loss": 5.882895085960627e-05,
|
| 1264 |
+
"step": 17800
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.6260901162790697,
|
| 1268 |
+
"grad_norm": 0.001823420519940555,
|
| 1269 |
+
"learning_rate": 9.350018168604652e-06,
|
| 1270 |
+
"loss": 0.0002583874017000198,
|
| 1271 |
+
"step": 17900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.635174418604651,
|
| 1275 |
+
"grad_norm": 0.6227073073387146,
|
| 1276 |
+
"learning_rate": 9.122910610465116e-06,
|
| 1277 |
+
"loss": 0.0007115737348794937,
|
| 1278 |
+
"step": 18000
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.6442587209302326,
|
| 1282 |
+
"grad_norm": 0.0017345056403428316,
|
| 1283 |
+
"learning_rate": 8.895803052325581e-06,
|
| 1284 |
+
"loss": 0.00027986690402030944,
|
| 1285 |
+
"step": 18100
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.653343023255814,
|
| 1289 |
+
"grad_norm": 0.0009033643291331828,
|
| 1290 |
+
"learning_rate": 8.668695494186047e-06,
|
| 1291 |
+
"loss": 0.0002979452162981033,
|
| 1292 |
+
"step": 18200
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.6624273255813953,
|
| 1296 |
+
"grad_norm": 0.00028923453646712005,
|
| 1297 |
+
"learning_rate": 8.441587936046511e-06,
|
| 1298 |
+
"loss": 1.926603843457997e-05,
|
| 1299 |
+
"step": 18300
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.6715116279069768,
|
| 1303 |
+
"grad_norm": 0.0006103311898186803,
|
| 1304 |
+
"learning_rate": 8.214480377906978e-06,
|
| 1305 |
+
"loss": 6.331724114716053e-05,
|
| 1306 |
+
"step": 18400
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.6805959302325582,
|
| 1310 |
+
"grad_norm": 0.0002818437642417848,
|
| 1311 |
+
"learning_rate": 7.987372819767442e-06,
|
| 1312 |
+
"loss": 0.00014105773530900478,
|
| 1313 |
+
"step": 18500
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.6896802325581395,
|
| 1317 |
+
"grad_norm": 0.0011020454112440348,
|
| 1318 |
+
"learning_rate": 7.760265261627907e-06,
|
| 1319 |
+
"loss": 0.0002292773686349392,
|
| 1320 |
+
"step": 18600
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.698764534883721,
|
| 1324 |
+
"grad_norm": 0.001500884653069079,
|
| 1325 |
+
"learning_rate": 7.533157703488372e-06,
|
| 1326 |
+
"loss": 0.00011514685116708279,
|
| 1327 |
+
"step": 18700
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1.7078488372093024,
|
| 1331 |
+
"grad_norm": 0.00016000888717826456,
|
| 1332 |
+
"learning_rate": 7.306050145348838e-06,
|
| 1333 |
+
"loss": 2.0523781422525645e-05,
|
| 1334 |
+
"step": 18800
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.7169331395348837,
|
| 1338 |
+
"grad_norm": 0.003748674876987934,
|
| 1339 |
+
"learning_rate": 7.078942587209303e-06,
|
| 1340 |
+
"loss": 0.00043839264661073686,
|
| 1341 |
+
"step": 18900
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.7260174418604652,
|
| 1345 |
+
"grad_norm": 0.06772974133491516,
|
| 1346 |
+
"learning_rate": 6.8518350290697685e-06,
|
| 1347 |
+
"loss": 2.126413397490978e-05,
|
| 1348 |
+
"step": 19000
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.7351017441860463,
|
| 1352 |
+
"grad_norm": 0.0004789210797753185,
|
| 1353 |
+
"learning_rate": 6.624727470930232e-06,
|
| 1354 |
+
"loss": 0.00020929597318172455,
|
| 1355 |
+
"step": 19100
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.744186046511628,
|
| 1359 |
+
"grad_norm": 7.055519381538033e-05,
|
| 1360 |
+
"learning_rate": 6.397619912790697e-06,
|
| 1361 |
+
"loss": 0.00032072752714157104,
|
| 1362 |
+
"step": 19200
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.7532703488372094,
|
| 1366 |
+
"grad_norm": 0.00010641128028510138,
|
| 1367 |
+
"learning_rate": 6.1705123546511635e-06,
|
| 1368 |
+
"loss": 8.132393471896649e-05,
|
| 1369 |
+
"step": 19300
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.7623546511627906,
|
| 1373 |
+
"grad_norm": 0.0005615473492071033,
|
| 1374 |
+
"learning_rate": 5.943404796511629e-06,
|
| 1375 |
+
"loss": 9.162256610579788e-06,
|
| 1376 |
+
"step": 19400
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.771438953488372,
|
| 1380 |
+
"grad_norm": 0.004505404736846685,
|
| 1381 |
+
"learning_rate": 5.716297238372093e-06,
|
| 1382 |
+
"loss": 9.514865465462207e-05,
|
| 1383 |
+
"step": 19500
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.7805232558139537,
|
| 1387 |
+
"grad_norm": 0.0011890050955116749,
|
| 1388 |
+
"learning_rate": 5.4891896802325586e-06,
|
| 1389 |
+
"loss": 3.7443286273628474e-05,
|
| 1390 |
+
"step": 19600
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.7896075581395348,
|
| 1394 |
+
"grad_norm": 0.000441042153397575,
|
| 1395 |
+
"learning_rate": 5.262082122093023e-06,
|
| 1396 |
+
"loss": 5.087008234113455e-05,
|
| 1397 |
+
"step": 19700
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.7986918604651163,
|
| 1401 |
+
"grad_norm": 0.0001777316356310621,
|
| 1402 |
+
"learning_rate": 5.034974563953489e-06,
|
| 1403 |
+
"loss": 0.00017192648723721505,
|
| 1404 |
+
"step": 19800
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.8077761627906976,
|
| 1408 |
+
"grad_norm": 0.00015439293929375708,
|
| 1409 |
+
"learning_rate": 4.8078670058139536e-06,
|
| 1410 |
+
"loss": 0.0001058769691735506,
|
| 1411 |
+
"step": 19900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 1.816860465116279,
|
| 1415 |
+
"grad_norm": 0.0016349812503904104,
|
| 1416 |
+
"learning_rate": 4.580759447674419e-06,
|
| 1417 |
+
"loss": 1.0620863176882267e-05,
|
| 1418 |
+
"step": 20000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.8259447674418605,
|
| 1422 |
+
"grad_norm": 9.556530858390033e-05,
|
| 1423 |
+
"learning_rate": 4.353651889534884e-06,
|
| 1424 |
+
"loss": 0.00019107908010482789,
|
| 1425 |
+
"step": 20100
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.8350290697674418,
|
| 1429 |
+
"grad_norm": 0.0002568990457803011,
|
| 1430 |
+
"learning_rate": 4.126544331395349e-06,
|
| 1431 |
+
"loss": 0.00026563439518213273,
|
| 1432 |
+
"step": 20200
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.8441133720930232,
|
| 1436 |
+
"grad_norm": 0.5719628930091858,
|
| 1437 |
+
"learning_rate": 3.899436773255814e-06,
|
| 1438 |
+
"loss": 5.403281655162573e-05,
|
| 1439 |
+
"step": 20300
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.8531976744186047,
|
| 1443 |
+
"grad_norm": 0.017383404076099396,
|
| 1444 |
+
"learning_rate": 3.672329215116279e-06,
|
| 1445 |
+
"loss": 0.00036652404814958573,
|
| 1446 |
+
"step": 20400
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.862281976744186,
|
| 1450 |
+
"grad_norm": 0.00022910887491889298,
|
| 1451 |
+
"learning_rate": 3.4452216569767445e-06,
|
| 1452 |
+
"loss": 0.000403200164437294,
|
| 1453 |
+
"step": 20500
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.8713662790697674,
|
| 1457 |
+
"grad_norm": 0.00033295468892902136,
|
| 1458 |
+
"learning_rate": 3.2181140988372097e-06,
|
| 1459 |
+
"loss": 8.646129630506039e-05,
|
| 1460 |
+
"step": 20600
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.880450581395349,
|
| 1464 |
+
"grad_norm": 0.00012523184705059975,
|
| 1465 |
+
"learning_rate": 2.9910065406976746e-06,
|
| 1466 |
+
"loss": 0.00016864996403455733,
|
| 1467 |
+
"step": 20700
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.8895348837209303,
|
| 1471 |
+
"grad_norm": 0.00022476979938801378,
|
| 1472 |
+
"learning_rate": 2.7638989825581395e-06,
|
| 1473 |
+
"loss": 5.167209077626467e-05,
|
| 1474 |
+
"step": 20800
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.8986191860465116,
|
| 1478 |
+
"grad_norm": 0.0002389108412899077,
|
| 1479 |
+
"learning_rate": 2.5367914244186048e-06,
|
| 1480 |
+
"loss": 6.591790355741977e-05,
|
| 1481 |
+
"step": 20900
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.9077034883720931,
|
| 1485 |
+
"grad_norm": 0.0013002109481021762,
|
| 1486 |
+
"learning_rate": 2.30968386627907e-06,
|
| 1487 |
+
"loss": 0.00010993644595146179,
|
| 1488 |
+
"step": 21000
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.9167877906976745,
|
| 1492 |
+
"grad_norm": 0.13188259303569794,
|
| 1493 |
+
"learning_rate": 2.082576308139535e-06,
|
| 1494 |
+
"loss": 0.00014082306995987892,
|
| 1495 |
+
"step": 21100
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.9258720930232558,
|
| 1499 |
+
"grad_norm": 0.0010737127158790827,
|
| 1500 |
+
"learning_rate": 1.85546875e-06,
|
| 1501 |
+
"loss": 0.00030819986015558244,
|
| 1502 |
+
"step": 21200
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.9349563953488373,
|
| 1506 |
+
"grad_norm": 7.047707185847685e-05,
|
| 1507 |
+
"learning_rate": 1.628361191860465e-06,
|
| 1508 |
+
"loss": 2.0567586179822683e-05,
|
| 1509 |
+
"step": 21300
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.9440406976744184,
|
| 1513 |
+
"grad_norm": 0.0001350079692201689,
|
| 1514 |
+
"learning_rate": 1.4012536337209304e-06,
|
| 1515 |
+
"loss": 7.989832083694636e-06,
|
| 1516 |
+
"step": 21400
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.953125,
|
| 1520 |
+
"grad_norm": 0.0008243785705417395,
|
| 1521 |
+
"learning_rate": 1.1741460755813954e-06,
|
| 1522 |
+
"loss": 0.0002570002153515816,
|
| 1523 |
+
"step": 21500
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.9622093023255816,
|
| 1527 |
+
"grad_norm": 0.001545518171042204,
|
| 1528 |
+
"learning_rate": 9.470385174418604e-07,
|
| 1529 |
+
"loss": 2.7221005875617266e-05,
|
| 1530 |
+
"step": 21600
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 1.9712936046511627,
|
| 1534 |
+
"grad_norm": 5.5686323321424425e-05,
|
| 1535 |
+
"learning_rate": 7.199309593023256e-07,
|
| 1536 |
+
"loss": 0.00010406752116978168,
|
| 1537 |
+
"step": 21700
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 1.9803779069767442,
|
| 1541 |
+
"grad_norm": 5.525942106032744e-05,
|
| 1542 |
+
"learning_rate": 4.928234011627908e-07,
|
| 1543 |
+
"loss": 3.853037022054195e-05,
|
| 1544 |
+
"step": 21800
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.9894622093023255,
|
| 1548 |
+
"grad_norm": 0.0002285480877617374,
|
| 1549 |
+
"learning_rate": 2.657158430232558e-07,
|
| 1550 |
+
"loss": 0.00048015639185905457,
|
| 1551 |
+
"step": 21900
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.9985465116279069,
|
| 1555 |
+
"grad_norm": 8.369733404833823e-05,
|
| 1556 |
+
"learning_rate": 3.860828488372093e-08,
|
| 1557 |
+
"loss": 6.105023785494268e-06,
|
| 1558 |
+
"step": 22000
|
| 1559 |
+
}
|
| 1560 |
+
],
|
| 1561 |
+
"logging_steps": 100,
|
| 1562 |
+
"max_steps": 22016,
|
| 1563 |
+
"num_input_tokens_seen": 0,
|
| 1564 |
+
"num_train_epochs": 2,
|
| 1565 |
+
"save_steps": 500,
|
| 1566 |
+
"stateful_callbacks": {
|
| 1567 |
+
"TrainerControl": {
|
| 1568 |
+
"args": {
|
| 1569 |
+
"should_epoch_stop": false,
|
| 1570 |
+
"should_evaluate": false,
|
| 1571 |
+
"should_log": false,
|
| 1572 |
+
"should_save": true,
|
| 1573 |
+
"should_training_stop": false
|
| 1574 |
+
},
|
| 1575 |
+
"attributes": {}
|
| 1576 |
+
}
|
| 1577 |
+
},
|
| 1578 |
+
"total_flos": 1.341374477893632e+16,
|
| 1579 |
+
"train_batch_size": 8,
|
| 1580 |
+
"trial_name": null,
|
| 1581 |
+
"trial_params": null
|
| 1582 |
+
}
|
models/dostoievsky_v1/checkpoint-22000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:edab2539a81b65c2a76c4c39bed219aa02358f44f3286429f83165442a1b53fa
|
| 3 |
+
size 5329
|
models/dostoievsky_v1/checkpoint-22016/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.1,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"BartForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.1,
|
| 10 |
+
"bos_token_id": 0,
|
| 11 |
+
"classif_dropout": 0.1,
|
| 12 |
+
"classifier_dropout": 0.0,
|
| 13 |
+
"d_model": 768,
|
| 14 |
+
"decoder_attention_heads": 12,
|
| 15 |
+
"decoder_ffn_dim": 3072,
|
| 16 |
+
"decoder_layerdrop": 0.0,
|
| 17 |
+
"decoder_layers": 6,
|
| 18 |
+
"decoder_start_token_id": 2,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 12,
|
| 22 |
+
"encoder_ffn_dim": 3072,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": true,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1,
|
| 39 |
+
"LABEL_2": 2
|
| 40 |
+
},
|
| 41 |
+
"max_position_embeddings": 1024,
|
| 42 |
+
"model_type": "bart",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": true,
|
| 45 |
+
"pad_token_id": 1,
|
| 46 |
+
"scale_embedding": false,
|
| 47 |
+
"task_specific_params": {
|
| 48 |
+
"summarization": {
|
| 49 |
+
"length_penalty": 1.0,
|
| 50 |
+
"max_length": 128,
|
| 51 |
+
"min_length": 12,
|
| 52 |
+
"num_beams": 4
|
| 53 |
+
},
|
| 54 |
+
"summarization_cnn": {
|
| 55 |
+
"length_penalty": 2.0,
|
| 56 |
+
"max_length": 142,
|
| 57 |
+
"min_length": 56,
|
| 58 |
+
"num_beams": 4
|
| 59 |
+
},
|
| 60 |
+
"summarization_xsum": {
|
| 61 |
+
"length_penalty": 1.0,
|
| 62 |
+
"max_length": 62,
|
| 63 |
+
"min_length": 11,
|
| 64 |
+
"num_beams": 6
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.5.0",
|
| 69 |
+
"use_cache": false,
|
| 70 |
+
"vocab_size": 50265
|
| 71 |
+
}
|
models/dostoievsky_v1/checkpoint-22016/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"early_stopping": true,
|
| 6 |
+
"eos_token_id": [
|
| 7 |
+
2
|
| 8 |
+
],
|
| 9 |
+
"forced_bos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 2,
|
| 11 |
+
"no_repeat_ngram_size": 3,
|
| 12 |
+
"num_beams": 4,
|
| 13 |
+
"pad_token_id": 1,
|
| 14 |
+
"transformers_version": "5.5.0",
|
| 15 |
+
"use_cache": true
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-22016/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d9dc9ffae85920773b2b5774f37b9630b197fc455008db011e6288a451870cb8
|
| 3 |
+
size 557912620
|
models/dostoievsky_v1/checkpoint-22016/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3cef9d187ecb037f2d7e4d910620d3c26ea5e61d28c9406a25018cdeb2412f2
|
| 3 |
+
size 1115583947
|
models/dostoievsky_v1/checkpoint-22016/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bef8c98d0370faf4d9e294ae6bb5a54459b7df1614ab0d628dfa9ab8c16113d
|
| 3 |
+
size 14645
|
models/dostoievsky_v1/checkpoint-22016/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de1988b91eb0266803c28f7f4b4c45861f95d52f7293e9c027e38a68aedd566a
|
| 3 |
+
size 1383
|
models/dostoievsky_v1/checkpoint-22016/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af2fefc736f21c2829d14de379d49371f2072daa593ac72eb437dc21a1f555df
|
| 3 |
+
size 1465
|
models/dostoievsky_v1/checkpoint-22016/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
models/dostoievsky_v1/checkpoint-22016/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"mask_token": "<mask>",
|
| 10 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"sep_token": "</s>",
|
| 13 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 14 |
+
"trim_offsets": true,
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
models/dostoievsky_v1/checkpoint-22016/trainer_state.json
ADDED
|
@@ -0,0 +1,1582 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 22016,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.009084302325581396,
|
| 14 |
+
"grad_norm": 1.9626818895339966,
|
| 15 |
+
"learning_rate": 4.9775163517441866e-05,
|
| 16 |
+
"loss": 3.2140347290039064,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.018168604651162792,
|
| 21 |
+
"grad_norm": 1.511451244354248,
|
| 22 |
+
"learning_rate": 4.9548055959302324e-05,
|
| 23 |
+
"loss": 0.01261173963546753,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.027252906976744186,
|
| 28 |
+
"grad_norm": 0.18194368481636047,
|
| 29 |
+
"learning_rate": 4.9320948401162795e-05,
|
| 30 |
+
"loss": 0.009183254837989808,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.036337209302325583,
|
| 35 |
+
"grad_norm": 0.27592894434928894,
|
| 36 |
+
"learning_rate": 4.909384084302326e-05,
|
| 37 |
+
"loss": 0.0027826914191246034,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.045421511627906974,
|
| 42 |
+
"grad_norm": 1.5584100484848022,
|
| 43 |
+
"learning_rate": 4.8866733284883724e-05,
|
| 44 |
+
"loss": 0.0029672542214393615,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05450581395348837,
|
| 49 |
+
"grad_norm": 0.02698471024632454,
|
| 50 |
+
"learning_rate": 4.863962572674419e-05,
|
| 51 |
+
"loss": 0.004296095669269561,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06359011627906977,
|
| 56 |
+
"grad_norm": 0.137993723154068,
|
| 57 |
+
"learning_rate": 4.841251816860465e-05,
|
| 58 |
+
"loss": 0.0031410756707191466,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07267441860465117,
|
| 63 |
+
"grad_norm": 0.07874622195959091,
|
| 64 |
+
"learning_rate": 4.818541061046512e-05,
|
| 65 |
+
"loss": 0.003037240505218506,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08175872093023256,
|
| 70 |
+
"grad_norm": 0.013660268858075142,
|
| 71 |
+
"learning_rate": 4.795830305232558e-05,
|
| 72 |
+
"loss": 0.0031082597374916076,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.09084302325581395,
|
| 77 |
+
"grad_norm": 0.3354911804199219,
|
| 78 |
+
"learning_rate": 4.7731195494186046e-05,
|
| 79 |
+
"loss": 0.0048285979032516475,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09992732558139535,
|
| 84 |
+
"grad_norm": 0.43667030334472656,
|
| 85 |
+
"learning_rate": 4.750408793604652e-05,
|
| 86 |
+
"loss": 0.0019270157814025878,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10901162790697674,
|
| 91 |
+
"grad_norm": 0.03849343582987785,
|
| 92 |
+
"learning_rate": 4.7276980377906975e-05,
|
| 93 |
+
"loss": 0.001011388897895813,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11809593023255814,
|
| 98 |
+
"grad_norm": 0.017985744401812553,
|
| 99 |
+
"learning_rate": 4.7049872819767446e-05,
|
| 100 |
+
"loss": 0.0018200889229774476,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12718023255813954,
|
| 105 |
+
"grad_norm": 0.05241613835096359,
|
| 106 |
+
"learning_rate": 4.682276526162791e-05,
|
| 107 |
+
"loss": 0.004675151705741882,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13626453488372092,
|
| 112 |
+
"grad_norm": 0.16119782626628876,
|
| 113 |
+
"learning_rate": 4.6595657703488375e-05,
|
| 114 |
+
"loss": 0.0025834646821022034,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.14534883720930233,
|
| 119 |
+
"grad_norm": 0.8985283374786377,
|
| 120 |
+
"learning_rate": 4.636855014534884e-05,
|
| 121 |
+
"loss": 0.001825849711894989,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.15443313953488372,
|
| 126 |
+
"grad_norm": 0.021285999566316605,
|
| 127 |
+
"learning_rate": 4.6141442587209304e-05,
|
| 128 |
+
"loss": 0.004164438843727112,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.16351744186046513,
|
| 133 |
+
"grad_norm": 0.01013926975429058,
|
| 134 |
+
"learning_rate": 4.591433502906977e-05,
|
| 135 |
+
"loss": 0.002582077383995056,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1726017441860465,
|
| 140 |
+
"grad_norm": 0.005713903810828924,
|
| 141 |
+
"learning_rate": 4.568722747093023e-05,
|
| 142 |
+
"loss": 0.0026018735766410827,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1816860465116279,
|
| 147 |
+
"grad_norm": 0.019032707437872887,
|
| 148 |
+
"learning_rate": 4.54601199127907e-05,
|
| 149 |
+
"loss": 0.0017240011692047118,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1907703488372093,
|
| 154 |
+
"grad_norm": 0.03191379830241203,
|
| 155 |
+
"learning_rate": 4.523301235465117e-05,
|
| 156 |
+
"loss": 0.0022749193012714386,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1998546511627907,
|
| 161 |
+
"grad_norm": 0.12945719063282013,
|
| 162 |
+
"learning_rate": 4.5005904796511626e-05,
|
| 163 |
+
"loss": 0.0017636509239673615,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2089389534883721,
|
| 168 |
+
"grad_norm": 0.01494428887963295,
|
| 169 |
+
"learning_rate": 4.47787972383721e-05,
|
| 170 |
+
"loss": 0.0017163331806659698,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2180232558139535,
|
| 175 |
+
"grad_norm": 0.08089974522590637,
|
| 176 |
+
"learning_rate": 4.455168968023256e-05,
|
| 177 |
+
"loss": 0.0023377402126789092,
|
| 178 |
+
"step": 2400
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.22710755813953487,
|
| 182 |
+
"grad_norm": 0.004136559087783098,
|
| 183 |
+
"learning_rate": 4.4324582122093026e-05,
|
| 184 |
+
"loss": 0.001353910267353058,
|
| 185 |
+
"step": 2500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.23619186046511628,
|
| 189 |
+
"grad_norm": 0.25111478567123413,
|
| 190 |
+
"learning_rate": 4.409747456395349e-05,
|
| 191 |
+
"loss": 0.0022939679026603697,
|
| 192 |
+
"step": 2600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.24527616279069767,
|
| 196 |
+
"grad_norm": 0.963623583316803,
|
| 197 |
+
"learning_rate": 4.3870367005813955e-05,
|
| 198 |
+
"loss": 0.0018738456070423125,
|
| 199 |
+
"step": 2700
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2543604651162791,
|
| 203 |
+
"grad_norm": 0.6530899405479431,
|
| 204 |
+
"learning_rate": 4.364325944767442e-05,
|
| 205 |
+
"loss": 0.0015843257308006287,
|
| 206 |
+
"step": 2800
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.26344476744186046,
|
| 210 |
+
"grad_norm": 0.0018010369967669249,
|
| 211 |
+
"learning_rate": 4.3416151889534884e-05,
|
| 212 |
+
"loss": 0.0006610354781150817,
|
| 213 |
+
"step": 2900
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.27252906976744184,
|
| 217 |
+
"grad_norm": 0.1644030064344406,
|
| 218 |
+
"learning_rate": 4.318904433139535e-05,
|
| 219 |
+
"loss": 0.0031197205185890197,
|
| 220 |
+
"step": 3000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.28161337209302323,
|
| 224 |
+
"grad_norm": 0.14440134167671204,
|
| 225 |
+
"learning_rate": 4.296193677325582e-05,
|
| 226 |
+
"loss": 0.0011694706976413728,
|
| 227 |
+
"step": 3100
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.29069767441860467,
|
| 231 |
+
"grad_norm": 0.0124755734577775,
|
| 232 |
+
"learning_rate": 4.273482921511628e-05,
|
| 233 |
+
"loss": 0.0013250903785228729,
|
| 234 |
+
"step": 3200
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.29978197674418605,
|
| 238 |
+
"grad_norm": 0.15910762548446655,
|
| 239 |
+
"learning_rate": 4.250772165697675e-05,
|
| 240 |
+
"loss": 0.0007842753082513809,
|
| 241 |
+
"step": 3300
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.30886627906976744,
|
| 245 |
+
"grad_norm": 0.011540662497282028,
|
| 246 |
+
"learning_rate": 4.228061409883721e-05,
|
| 247 |
+
"loss": 0.0018773898482322693,
|
| 248 |
+
"step": 3400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3179505813953488,
|
| 252 |
+
"grad_norm": 0.005148735363036394,
|
| 253 |
+
"learning_rate": 4.205350654069768e-05,
|
| 254 |
+
"loss": 0.001056775525212288,
|
| 255 |
+
"step": 3500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.32703488372093026,
|
| 259 |
+
"grad_norm": 0.007442768197506666,
|
| 260 |
+
"learning_rate": 4.182639898255814e-05,
|
| 261 |
+
"loss": 0.0014348265528678894,
|
| 262 |
+
"step": 3600
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.33611918604651164,
|
| 266 |
+
"grad_norm": 0.00586783979088068,
|
| 267 |
+
"learning_rate": 4.1599291424418606e-05,
|
| 268 |
+
"loss": 0.0014780126512050629,
|
| 269 |
+
"step": 3700
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.345203488372093,
|
| 273 |
+
"grad_norm": 0.006909696385264397,
|
| 274 |
+
"learning_rate": 4.137218386627907e-05,
|
| 275 |
+
"loss": 0.0019674110412597656,
|
| 276 |
+
"step": 3800
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3542877906976744,
|
| 280 |
+
"grad_norm": 0.020171863958239555,
|
| 281 |
+
"learning_rate": 4.1145076308139535e-05,
|
| 282 |
+
"loss": 0.00249418705701828,
|
| 283 |
+
"step": 3900
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3633720930232558,
|
| 287 |
+
"grad_norm": 0.031723715364933014,
|
| 288 |
+
"learning_rate": 4.091796875e-05,
|
| 289 |
+
"loss": 0.0030515387654304503,
|
| 290 |
+
"step": 4000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.37245639534883723,
|
| 294 |
+
"grad_norm": 0.30729565024375916,
|
| 295 |
+
"learning_rate": 4.069086119186047e-05,
|
| 296 |
+
"loss": 0.0009509802609682083,
|
| 297 |
+
"step": 4100
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3815406976744186,
|
| 301 |
+
"grad_norm": 0.0021714100148528814,
|
| 302 |
+
"learning_rate": 4.046375363372093e-05,
|
| 303 |
+
"loss": 0.0013920699059963226,
|
| 304 |
+
"step": 4200
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.390625,
|
| 308 |
+
"grad_norm": 0.001017636270262301,
|
| 309 |
+
"learning_rate": 4.02366460755814e-05,
|
| 310 |
+
"loss": 0.0010310819000005722,
|
| 311 |
+
"step": 4300
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3997093023255814,
|
| 315 |
+
"grad_norm": 0.0009663284290581942,
|
| 316 |
+
"learning_rate": 4.0009538517441864e-05,
|
| 317 |
+
"loss": 0.001098722442984581,
|
| 318 |
+
"step": 4400
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.40879360465116277,
|
| 322 |
+
"grad_norm": 0.003507931251078844,
|
| 323 |
+
"learning_rate": 3.978243095930232e-05,
|
| 324 |
+
"loss": 0.0013840131461620331,
|
| 325 |
+
"step": 4500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4178779069767442,
|
| 329 |
+
"grad_norm": 0.21988199651241302,
|
| 330 |
+
"learning_rate": 3.955532340116279e-05,
|
| 331 |
+
"loss": 0.0017136169970035553,
|
| 332 |
+
"step": 4600
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4269622093023256,
|
| 336 |
+
"grad_norm": 0.0012972657568752766,
|
| 337 |
+
"learning_rate": 3.932821584302326e-05,
|
| 338 |
+
"loss": 0.0008040449023246765,
|
| 339 |
+
"step": 4700
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.436046511627907,
|
| 343 |
+
"grad_norm": 0.033857911825180054,
|
| 344 |
+
"learning_rate": 3.910110828488372e-05,
|
| 345 |
+
"loss": 0.0005662231892347336,
|
| 346 |
+
"step": 4800
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.44513081395348836,
|
| 350 |
+
"grad_norm": 0.0008128180052153766,
|
| 351 |
+
"learning_rate": 3.8874000726744187e-05,
|
| 352 |
+
"loss": 0.001040520742535591,
|
| 353 |
+
"step": 4900
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.45421511627906974,
|
| 357 |
+
"grad_norm": 0.0017111338675022125,
|
| 358 |
+
"learning_rate": 3.864689316860465e-05,
|
| 359 |
+
"loss": 0.0010922805964946748,
|
| 360 |
+
"step": 5000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.4632994186046512,
|
| 364 |
+
"grad_norm": 0.0013291583163663745,
|
| 365 |
+
"learning_rate": 3.841978561046512e-05,
|
| 366 |
+
"loss": 0.0005277743935585021,
|
| 367 |
+
"step": 5100
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.47238372093023256,
|
| 371 |
+
"grad_norm": 0.09846807271242142,
|
| 372 |
+
"learning_rate": 3.819267805232558e-05,
|
| 373 |
+
"loss": 0.0010997675359249114,
|
| 374 |
+
"step": 5200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.48146802325581395,
|
| 378 |
+
"grad_norm": 0.005185174290090799,
|
| 379 |
+
"learning_rate": 3.796557049418605e-05,
|
| 380 |
+
"loss": 0.000698111355304718,
|
| 381 |
+
"step": 5300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.49055232558139533,
|
| 385 |
+
"grad_norm": 0.019086388871073723,
|
| 386 |
+
"learning_rate": 3.7738462936046515e-05,
|
| 387 |
+
"loss": 0.0009016367793083191,
|
| 388 |
+
"step": 5400
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.49963662790697677,
|
| 392 |
+
"grad_norm": 0.05122831463813782,
|
| 393 |
+
"learning_rate": 3.751135537790697e-05,
|
| 394 |
+
"loss": 0.0015524370968341826,
|
| 395 |
+
"step": 5500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5087209302325582,
|
| 399 |
+
"grad_norm": 0.012596211396157742,
|
| 400 |
+
"learning_rate": 3.7284247819767444e-05,
|
| 401 |
+
"loss": 0.00220917209982872,
|
| 402 |
+
"step": 5600
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5178052325581395,
|
| 406 |
+
"grad_norm": 0.009361029602587223,
|
| 407 |
+
"learning_rate": 3.705714026162791e-05,
|
| 408 |
+
"loss": 0.0010362663865089417,
|
| 409 |
+
"step": 5700
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5268895348837209,
|
| 413 |
+
"grad_norm": 0.5205378532409668,
|
| 414 |
+
"learning_rate": 3.683003270348837e-05,
|
| 415 |
+
"loss": 0.0022550773620605468,
|
| 416 |
+
"step": 5800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5359738372093024,
|
| 420 |
+
"grad_norm": 0.0006470124353654683,
|
| 421 |
+
"learning_rate": 3.660292514534884e-05,
|
| 422 |
+
"loss": 0.0011264414340257645,
|
| 423 |
+
"step": 5900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5450581395348837,
|
| 427 |
+
"grad_norm": 0.058231666684150696,
|
| 428 |
+
"learning_rate": 3.63758175872093e-05,
|
| 429 |
+
"loss": 0.0011266635358333588,
|
| 430 |
+
"step": 6000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5541424418604651,
|
| 434 |
+
"grad_norm": 0.0013441125629469752,
|
| 435 |
+
"learning_rate": 3.614871002906977e-05,
|
| 436 |
+
"loss": 0.0008001205325126648,
|
| 437 |
+
"step": 6100
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5632267441860465,
|
| 441 |
+
"grad_norm": 0.0021539030130952597,
|
| 442 |
+
"learning_rate": 3.592160247093023e-05,
|
| 443 |
+
"loss": 0.0008849448710680007,
|
| 444 |
+
"step": 6200
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5723110465116279,
|
| 448 |
+
"grad_norm": 0.9019960761070251,
|
| 449 |
+
"learning_rate": 3.56944949127907e-05,
|
| 450 |
+
"loss": 0.0006517694145441055,
|
| 451 |
+
"step": 6300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5813953488372093,
|
| 455 |
+
"grad_norm": 0.001283760997466743,
|
| 456 |
+
"learning_rate": 3.546738735465117e-05,
|
| 457 |
+
"loss": 0.0021727623045444486,
|
| 458 |
+
"step": 6400
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5904796511627907,
|
| 462 |
+
"grad_norm": 0.005666423588991165,
|
| 463 |
+
"learning_rate": 3.5240279796511624e-05,
|
| 464 |
+
"loss": 0.0012587083876132966,
|
| 465 |
+
"step": 6500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5995639534883721,
|
| 469 |
+
"grad_norm": 1.4842944145202637,
|
| 470 |
+
"learning_rate": 3.5013172238372096e-05,
|
| 471 |
+
"loss": 0.001180112287402153,
|
| 472 |
+
"step": 6600
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6086482558139535,
|
| 476 |
+
"grad_norm": 0.02973734401166439,
|
| 477 |
+
"learning_rate": 3.478606468023256e-05,
|
| 478 |
+
"loss": 0.0010163599252700805,
|
| 479 |
+
"step": 6700
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6177325581395349,
|
| 483 |
+
"grad_norm": 0.00399527233093977,
|
| 484 |
+
"learning_rate": 3.4558957122093024e-05,
|
| 485 |
+
"loss": 0.0010414445400238038,
|
| 486 |
+
"step": 6800
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6268168604651163,
|
| 490 |
+
"grad_norm": 0.39672717452049255,
|
| 491 |
+
"learning_rate": 3.433184956395349e-05,
|
| 492 |
+
"loss": 0.0012094499170780182,
|
| 493 |
+
"step": 6900
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6359011627906976,
|
| 497 |
+
"grad_norm": 0.5308993458747864,
|
| 498 |
+
"learning_rate": 3.410474200581395e-05,
|
| 499 |
+
"loss": 0.0006021633744239807,
|
| 500 |
+
"step": 7000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6449854651162791,
|
| 504 |
+
"grad_norm": 0.003343795659020543,
|
| 505 |
+
"learning_rate": 3.3877634447674425e-05,
|
| 506 |
+
"loss": 0.0005116893351078033,
|
| 507 |
+
"step": 7100
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6540697674418605,
|
| 511 |
+
"grad_norm": 0.0033038391266018152,
|
| 512 |
+
"learning_rate": 3.365052688953488e-05,
|
| 513 |
+
"loss": 0.002106922417879105,
|
| 514 |
+
"step": 7200
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6631540697674418,
|
| 518 |
+
"grad_norm": 0.004004980903118849,
|
| 519 |
+
"learning_rate": 3.3423419331395353e-05,
|
| 520 |
+
"loss": 0.0010233993828296661,
|
| 521 |
+
"step": 7300
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6722383720930233,
|
| 525 |
+
"grad_norm": 0.00490298168733716,
|
| 526 |
+
"learning_rate": 3.319631177325582e-05,
|
| 527 |
+
"loss": 0.0007062336057424546,
|
| 528 |
+
"step": 7400
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6813226744186046,
|
| 532 |
+
"grad_norm": 0.04243200644850731,
|
| 533 |
+
"learning_rate": 3.2969204215116276e-05,
|
| 534 |
+
"loss": 0.0005479569733142852,
|
| 535 |
+
"step": 7500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.690406976744186,
|
| 539 |
+
"grad_norm": 0.008556894026696682,
|
| 540 |
+
"learning_rate": 3.274209665697675e-05,
|
| 541 |
+
"loss": 0.0010938134789466858,
|
| 542 |
+
"step": 7600
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6994912790697675,
|
| 546 |
+
"grad_norm": 0.016938723623752594,
|
| 547 |
+
"learning_rate": 3.251498909883721e-05,
|
| 548 |
+
"loss": 0.0009466408193111419,
|
| 549 |
+
"step": 7700
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7085755813953488,
|
| 553 |
+
"grad_norm": 0.09418574720621109,
|
| 554 |
+
"learning_rate": 3.2287881540697676e-05,
|
| 555 |
+
"loss": 0.00145876482129097,
|
| 556 |
+
"step": 7800
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7176598837209303,
|
| 560 |
+
"grad_norm": 0.002359782112762332,
|
| 561 |
+
"learning_rate": 3.206077398255814e-05,
|
| 562 |
+
"loss": 0.0015742655098438263,
|
| 563 |
+
"step": 7900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7267441860465116,
|
| 567 |
+
"grad_norm": 0.3374776840209961,
|
| 568 |
+
"learning_rate": 3.1833666424418605e-05,
|
| 569 |
+
"loss": 0.0007466593384742737,
|
| 570 |
+
"step": 8000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.735828488372093,
|
| 574 |
+
"grad_norm": 0.000735403154976666,
|
| 575 |
+
"learning_rate": 3.1606558866279076e-05,
|
| 576 |
+
"loss": 0.0008171546459197998,
|
| 577 |
+
"step": 8100
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7449127906976745,
|
| 581 |
+
"grad_norm": 0.01963644102215767,
|
| 582 |
+
"learning_rate": 3.1379451308139533e-05,
|
| 583 |
+
"loss": 0.0007363802939653396,
|
| 584 |
+
"step": 8200
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7539970930232558,
|
| 588 |
+
"grad_norm": 0.09964141249656677,
|
| 589 |
+
"learning_rate": 3.1152343750000005e-05,
|
| 590 |
+
"loss": 0.00030826406553387644,
|
| 591 |
+
"step": 8300
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7630813953488372,
|
| 595 |
+
"grad_norm": 0.0029934593476355076,
|
| 596 |
+
"learning_rate": 3.092523619186047e-05,
|
| 597 |
+
"loss": 0.0003284827247262001,
|
| 598 |
+
"step": 8400
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7721656976744186,
|
| 602 |
+
"grad_norm": 0.002162993187084794,
|
| 603 |
+
"learning_rate": 3.069812863372093e-05,
|
| 604 |
+
"loss": 0.0010288888961076737,
|
| 605 |
+
"step": 8500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.78125,
|
| 609 |
+
"grad_norm": 0.013634397648274899,
|
| 610 |
+
"learning_rate": 3.0471021075581398e-05,
|
| 611 |
+
"loss": 0.0006983913481235504,
|
| 612 |
+
"step": 8600
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7903343023255814,
|
| 616 |
+
"grad_norm": 0.0833001658320427,
|
| 617 |
+
"learning_rate": 3.0243913517441862e-05,
|
| 618 |
+
"loss": 0.0009967025369405746,
|
| 619 |
+
"step": 8700
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7994186046511628,
|
| 623 |
+
"grad_norm": 0.4697379767894745,
|
| 624 |
+
"learning_rate": 3.0016805959302323e-05,
|
| 625 |
+
"loss": 0.0007716407626867294,
|
| 626 |
+
"step": 8800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8085029069767442,
|
| 630 |
+
"grad_norm": 0.04588587209582329,
|
| 631 |
+
"learning_rate": 2.978969840116279e-05,
|
| 632 |
+
"loss": 0.000263803955167532,
|
| 633 |
+
"step": 8900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8175872093023255,
|
| 637 |
+
"grad_norm": 0.003420337103307247,
|
| 638 |
+
"learning_rate": 2.956259084302326e-05,
|
| 639 |
+
"loss": 0.0009599439054727555,
|
| 640 |
+
"step": 9000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.826671511627907,
|
| 644 |
+
"grad_norm": 0.0024436134845018387,
|
| 645 |
+
"learning_rate": 2.9335483284883724e-05,
|
| 646 |
+
"loss": 0.012270723581314086,
|
| 647 |
+
"step": 9100
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8357558139534884,
|
| 651 |
+
"grad_norm": 0.001418731757439673,
|
| 652 |
+
"learning_rate": 2.9108375726744185e-05,
|
| 653 |
+
"loss": 0.0009713788330554962,
|
| 654 |
+
"step": 9200
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8448401162790697,
|
| 658 |
+
"grad_norm": 0.0026375153101980686,
|
| 659 |
+
"learning_rate": 2.8881268168604652e-05,
|
| 660 |
+
"loss": 0.0007654589414596558,
|
| 661 |
+
"step": 9300
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8539244186046512,
|
| 665 |
+
"grad_norm": 0.0013760777655988932,
|
| 666 |
+
"learning_rate": 2.865416061046512e-05,
|
| 667 |
+
"loss": 0.0006276721507310867,
|
| 668 |
+
"step": 9400
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8630087209302325,
|
| 672 |
+
"grad_norm": 0.0033582421019673347,
|
| 673 |
+
"learning_rate": 2.842705305232558e-05,
|
| 674 |
+
"loss": 0.0003520375117659569,
|
| 675 |
+
"step": 9500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.872093023255814,
|
| 679 |
+
"grad_norm": 0.0043012769892811775,
|
| 680 |
+
"learning_rate": 2.819994549418605e-05,
|
| 681 |
+
"loss": 0.0011182524263858794,
|
| 682 |
+
"step": 9600
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8811773255813954,
|
| 686 |
+
"grad_norm": 0.0015025343745946884,
|
| 687 |
+
"learning_rate": 2.7972837936046514e-05,
|
| 688 |
+
"loss": 0.00044553544372320173,
|
| 689 |
+
"step": 9700
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8902616279069767,
|
| 693 |
+
"grad_norm": 0.0018241156358271837,
|
| 694 |
+
"learning_rate": 2.7745730377906975e-05,
|
| 695 |
+
"loss": 0.00023134740069508553,
|
| 696 |
+
"step": 9800
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8993459302325582,
|
| 700 |
+
"grad_norm": 0.001082456554286182,
|
| 701 |
+
"learning_rate": 2.7518622819767442e-05,
|
| 702 |
+
"loss": 0.001018296480178833,
|
| 703 |
+
"step": 9900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.9084302325581395,
|
| 707 |
+
"grad_norm": 0.01155087724328041,
|
| 708 |
+
"learning_rate": 2.729151526162791e-05,
|
| 709 |
+
"loss": 0.0005589094385504722,
|
| 710 |
+
"step": 10000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.9175145348837209,
|
| 714 |
+
"grad_norm": 0.0022482366766780615,
|
| 715 |
+
"learning_rate": 2.7064407703488375e-05,
|
| 716 |
+
"loss": 0.000512191392481327,
|
| 717 |
+
"step": 10100
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9265988372093024,
|
| 721 |
+
"grad_norm": 0.20549768209457397,
|
| 722 |
+
"learning_rate": 2.6837300145348836e-05,
|
| 723 |
+
"loss": 0.00038118865340948106,
|
| 724 |
+
"step": 10200
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9356831395348837,
|
| 728 |
+
"grad_norm": 0.0013188497396185994,
|
| 729 |
+
"learning_rate": 2.6610192587209304e-05,
|
| 730 |
+
"loss": 0.0003574254736304283,
|
| 731 |
+
"step": 10300
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9447674418604651,
|
| 735 |
+
"grad_norm": 0.0006103936466388404,
|
| 736 |
+
"learning_rate": 2.638308502906977e-05,
|
| 737 |
+
"loss": 0.00024356411769986154,
|
| 738 |
+
"step": 10400
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9538517441860465,
|
| 742 |
+
"grad_norm": 0.00047453015577048063,
|
| 743 |
+
"learning_rate": 2.6155977470930233e-05,
|
| 744 |
+
"loss": 0.0013259868323802948,
|
| 745 |
+
"step": 10500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9629360465116279,
|
| 749 |
+
"grad_norm": 0.025760261341929436,
|
| 750 |
+
"learning_rate": 2.59288699127907e-05,
|
| 751 |
+
"loss": 0.0006390263140201569,
|
| 752 |
+
"step": 10600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9720203488372093,
|
| 756 |
+
"grad_norm": 0.0025282115675508976,
|
| 757 |
+
"learning_rate": 2.5701762354651165e-05,
|
| 758 |
+
"loss": 0.001759422868490219,
|
| 759 |
+
"step": 10700
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.9811046511627907,
|
| 763 |
+
"grad_norm": 2.990682601928711,
|
| 764 |
+
"learning_rate": 2.5474654796511626e-05,
|
| 765 |
+
"loss": 0.00041438560932874677,
|
| 766 |
+
"step": 10800
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9901889534883721,
|
| 770 |
+
"grad_norm": 0.0036354295443743467,
|
| 771 |
+
"learning_rate": 2.5247547238372094e-05,
|
| 772 |
+
"loss": 0.000467718243598938,
|
| 773 |
+
"step": 10900
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9992732558139535,
|
| 777 |
+
"grad_norm": 0.02648981101810932,
|
| 778 |
+
"learning_rate": 2.502043968023256e-05,
|
| 779 |
+
"loss": 0.0007214382290840149,
|
| 780 |
+
"step": 11000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0,
|
| 784 |
+
"eval_loss": 8.017317304620519e-05,
|
| 785 |
+
"eval_runtime": 132.6261,
|
| 786 |
+
"eval_samples_per_second": 165.993,
|
| 787 |
+
"eval_steps_per_second": 20.75,
|
| 788 |
+
"step": 11008
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.008357558139535,
|
| 792 |
+
"grad_norm": 0.0011333210859447718,
|
| 793 |
+
"learning_rate": 2.4793332122093026e-05,
|
| 794 |
+
"loss": 0.000609114095568657,
|
| 795 |
+
"step": 11100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.0174418604651163,
|
| 799 |
+
"grad_norm": 0.0006669509457424283,
|
| 800 |
+
"learning_rate": 2.4566224563953487e-05,
|
| 801 |
+
"loss": 0.000269341841340065,
|
| 802 |
+
"step": 11200
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.0265261627906976,
|
| 806 |
+
"grad_norm": 0.0006085751811042428,
|
| 807 |
+
"learning_rate": 2.4339117005813955e-05,
|
| 808 |
+
"loss": 0.00018008088693022728,
|
| 809 |
+
"step": 11300
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.035610465116279,
|
| 813 |
+
"grad_norm": 0.00029874974279664457,
|
| 814 |
+
"learning_rate": 2.411200944767442e-05,
|
| 815 |
+
"loss": 8.143479935824871e-05,
|
| 816 |
+
"step": 11400
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.0446947674418605,
|
| 820 |
+
"grad_norm": 0.0020065035205334425,
|
| 821 |
+
"learning_rate": 2.3884901889534887e-05,
|
| 822 |
+
"loss": 0.0004405264183878899,
|
| 823 |
+
"step": 11500
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.0537790697674418,
|
| 827 |
+
"grad_norm": 0.3139957785606384,
|
| 828 |
+
"learning_rate": 2.3657794331395348e-05,
|
| 829 |
+
"loss": 0.0007688279449939728,
|
| 830 |
+
"step": 11600
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.0628633720930232,
|
| 834 |
+
"grad_norm": 0.014869201928377151,
|
| 835 |
+
"learning_rate": 2.3430686773255813e-05,
|
| 836 |
+
"loss": 0.00045145414769649507,
|
| 837 |
+
"step": 11700
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.0719476744186047,
|
| 841 |
+
"grad_norm": 0.03592238947749138,
|
| 842 |
+
"learning_rate": 2.320357921511628e-05,
|
| 843 |
+
"loss": 0.0013660797476768493,
|
| 844 |
+
"step": 11800
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.081031976744186,
|
| 848 |
+
"grad_norm": 0.010270297527313232,
|
| 849 |
+
"learning_rate": 2.2976471656976745e-05,
|
| 850 |
+
"loss": 0.0005171676725149155,
|
| 851 |
+
"step": 11900
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.0901162790697674,
|
| 855 |
+
"grad_norm": 0.0026885548140853643,
|
| 856 |
+
"learning_rate": 2.2749364098837213e-05,
|
| 857 |
+
"loss": 0.00046057451516389846,
|
| 858 |
+
"step": 12000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.099200581395349,
|
| 862 |
+
"grad_norm": 0.005681134294718504,
|
| 863 |
+
"learning_rate": 2.2522256540697674e-05,
|
| 864 |
+
"loss": 0.0006235280632972718,
|
| 865 |
+
"step": 12100
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.1082848837209303,
|
| 869 |
+
"grad_norm": 0.0028001824393868446,
|
| 870 |
+
"learning_rate": 2.2295148982558138e-05,
|
| 871 |
+
"loss": 0.0008198145776987075,
|
| 872 |
+
"step": 12200
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.1173691860465116,
|
| 876 |
+
"grad_norm": 0.0006385694723576307,
|
| 877 |
+
"learning_rate": 2.2068041424418606e-05,
|
| 878 |
+
"loss": 0.0003722207620739937,
|
| 879 |
+
"step": 12300
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.1264534883720931,
|
| 883 |
+
"grad_norm": 0.0008649047813378274,
|
| 884 |
+
"learning_rate": 2.184093386627907e-05,
|
| 885 |
+
"loss": 0.0003394853696227074,
|
| 886 |
+
"step": 12400
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.1355377906976745,
|
| 890 |
+
"grad_norm": 0.004796006251126528,
|
| 891 |
+
"learning_rate": 2.1613826308139538e-05,
|
| 892 |
+
"loss": 0.00042502928525209424,
|
| 893 |
+
"step": 12500
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.1446220930232558,
|
| 897 |
+
"grad_norm": 0.001498043886385858,
|
| 898 |
+
"learning_rate": 2.138671875e-05,
|
| 899 |
+
"loss": 0.0009064202010631562,
|
| 900 |
+
"step": 12600
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.1537063953488371,
|
| 904 |
+
"grad_norm": 0.01664622500538826,
|
| 905 |
+
"learning_rate": 2.1159611191860464e-05,
|
| 906 |
+
"loss": 0.0002878718450665474,
|
| 907 |
+
"step": 12700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.1627906976744187,
|
| 911 |
+
"grad_norm": 0.010639426298439503,
|
| 912 |
+
"learning_rate": 2.093250363372093e-05,
|
| 913 |
+
"loss": 0.0011860337108373642,
|
| 914 |
+
"step": 12800
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.171875,
|
| 918 |
+
"grad_norm": 0.050644513219594955,
|
| 919 |
+
"learning_rate": 2.0705396075581396e-05,
|
| 920 |
+
"loss": 0.0006260576844215393,
|
| 921 |
+
"step": 12900
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.1809593023255813,
|
| 925 |
+
"grad_norm": 0.001398528809659183,
|
| 926 |
+
"learning_rate": 2.0478288517441864e-05,
|
| 927 |
+
"loss": 0.0001833665184676647,
|
| 928 |
+
"step": 13000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.1900436046511629,
|
| 932 |
+
"grad_norm": 0.0057320562191307545,
|
| 933 |
+
"learning_rate": 2.0251180959302325e-05,
|
| 934 |
+
"loss": 0.00032943256199359896,
|
| 935 |
+
"step": 13100
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.1991279069767442,
|
| 939 |
+
"grad_norm": 0.0002997924748342484,
|
| 940 |
+
"learning_rate": 2.002407340116279e-05,
|
| 941 |
+
"loss": 6.249572150409222e-05,
|
| 942 |
+
"step": 13200
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.2082122093023255,
|
| 946 |
+
"grad_norm": 0.0008130258647724986,
|
| 947 |
+
"learning_rate": 1.9796965843023257e-05,
|
| 948 |
+
"loss": 0.00026577839627861977,
|
| 949 |
+
"step": 13300
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.2172965116279069,
|
| 953 |
+
"grad_norm": 0.0006040658336132765,
|
| 954 |
+
"learning_rate": 1.956985828488372e-05,
|
| 955 |
+
"loss": 0.00023248378187417984,
|
| 956 |
+
"step": 13400
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.2263808139534884,
|
| 960 |
+
"grad_norm": 0.000803643895778805,
|
| 961 |
+
"learning_rate": 1.934275072674419e-05,
|
| 962 |
+
"loss": 0.0002730824239552021,
|
| 963 |
+
"step": 13500
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.2354651162790697,
|
| 967 |
+
"grad_norm": 0.0005482266424223781,
|
| 968 |
+
"learning_rate": 1.911564316860465e-05,
|
| 969 |
+
"loss": 0.0005848826467990875,
|
| 970 |
+
"step": 13600
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.244549418604651,
|
| 974 |
+
"grad_norm": 0.0034759771078824997,
|
| 975 |
+
"learning_rate": 1.8888535610465115e-05,
|
| 976 |
+
"loss": 0.0005462893471121788,
|
| 977 |
+
"step": 13700
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.2536337209302326,
|
| 981 |
+
"grad_norm": 0.0002226918877568096,
|
| 982 |
+
"learning_rate": 1.8661428052325583e-05,
|
| 983 |
+
"loss": 7.945694960653782e-05,
|
| 984 |
+
"step": 13800
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.262718023255814,
|
| 988 |
+
"grad_norm": 0.002374310279265046,
|
| 989 |
+
"learning_rate": 1.8434320494186047e-05,
|
| 990 |
+
"loss": 0.000744214728474617,
|
| 991 |
+
"step": 13900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.2718023255813953,
|
| 995 |
+
"grad_norm": 0.0009076696587726474,
|
| 996 |
+
"learning_rate": 1.8207212936046515e-05,
|
| 997 |
+
"loss": 0.0003250580281019211,
|
| 998 |
+
"step": 14000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.2808866279069768,
|
| 1002 |
+
"grad_norm": 0.010181100107729435,
|
| 1003 |
+
"learning_rate": 1.7980105377906976e-05,
|
| 1004 |
+
"loss": 0.0005765938758850097,
|
| 1005 |
+
"step": 14100
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.2899709302325582,
|
| 1009 |
+
"grad_norm": 0.0012442917795851827,
|
| 1010 |
+
"learning_rate": 1.775299781976744e-05,
|
| 1011 |
+
"loss": 0.00032320961356163024,
|
| 1012 |
+
"step": 14200
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.2990552325581395,
|
| 1016 |
+
"grad_norm": 0.0005617383285425603,
|
| 1017 |
+
"learning_rate": 1.752589026162791e-05,
|
| 1018 |
+
"loss": 0.0001776321791112423,
|
| 1019 |
+
"step": 14300
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.308139534883721,
|
| 1023 |
+
"grad_norm": 0.0010064981179311872,
|
| 1024 |
+
"learning_rate": 1.7298782703488373e-05,
|
| 1025 |
+
"loss": 0.00012774170376360415,
|
| 1026 |
+
"step": 14400
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.3172238372093024,
|
| 1030 |
+
"grad_norm": 0.00039220438338816166,
|
| 1031 |
+
"learning_rate": 1.707167514534884e-05,
|
| 1032 |
+
"loss": 0.0006190959364175796,
|
| 1033 |
+
"step": 14500
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.3263081395348837,
|
| 1037 |
+
"grad_norm": 0.0009137202869169414,
|
| 1038 |
+
"learning_rate": 1.6844567587209302e-05,
|
| 1039 |
+
"loss": 0.0007051125913858414,
|
| 1040 |
+
"step": 14600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.3353924418604652,
|
| 1044 |
+
"grad_norm": 0.0006007241318002343,
|
| 1045 |
+
"learning_rate": 1.6617460029069766e-05,
|
| 1046 |
+
"loss": 7.074063178151846e-05,
|
| 1047 |
+
"step": 14700
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.3444767441860466,
|
| 1051 |
+
"grad_norm": 0.0006039150175638497,
|
| 1052 |
+
"learning_rate": 1.6390352470930234e-05,
|
| 1053 |
+
"loss": 0.000489293597638607,
|
| 1054 |
+
"step": 14800
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.353561046511628,
|
| 1058 |
+
"grad_norm": 0.0014955669175833464,
|
| 1059 |
+
"learning_rate": 1.61632449127907e-05,
|
| 1060 |
+
"loss": 0.0004155828058719635,
|
| 1061 |
+
"step": 14900
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.3626453488372092,
|
| 1065 |
+
"grad_norm": 0.1508745402097702,
|
| 1066 |
+
"learning_rate": 1.5936137354651163e-05,
|
| 1067 |
+
"loss": 0.00024092141538858413,
|
| 1068 |
+
"step": 15000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.3717296511627908,
|
| 1072 |
+
"grad_norm": 0.0008113393560051918,
|
| 1073 |
+
"learning_rate": 1.5709029796511627e-05,
|
| 1074 |
+
"loss": 0.00040562458336353304,
|
| 1075 |
+
"step": 15100
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.380813953488372,
|
| 1079 |
+
"grad_norm": 0.0026459796354174614,
|
| 1080 |
+
"learning_rate": 1.5481922238372092e-05,
|
| 1081 |
+
"loss": 0.0007934534549713135,
|
| 1082 |
+
"step": 15200
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.3898982558139534,
|
| 1086 |
+
"grad_norm": 0.0005005365237593651,
|
| 1087 |
+
"learning_rate": 1.525481468023256e-05,
|
| 1088 |
+
"loss": 0.00023073634132742883,
|
| 1089 |
+
"step": 15300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.3989825581395348,
|
| 1093 |
+
"grad_norm": 0.02727348543703556,
|
| 1094 |
+
"learning_rate": 1.5027707122093024e-05,
|
| 1095 |
+
"loss": 0.0003524136170744896,
|
| 1096 |
+
"step": 15400
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.4080668604651163,
|
| 1100 |
+
"grad_norm": 0.0002921252744272351,
|
| 1101 |
+
"learning_rate": 1.4800599563953488e-05,
|
| 1102 |
+
"loss": 0.000525415763258934,
|
| 1103 |
+
"step": 15500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.4171511627906976,
|
| 1107 |
+
"grad_norm": 0.0016339289722964168,
|
| 1108 |
+
"learning_rate": 1.4573492005813955e-05,
|
| 1109 |
+
"loss": 0.0003766526654362678,
|
| 1110 |
+
"step": 15600
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.426235465116279,
|
| 1114 |
+
"grad_norm": 0.003485665889456868,
|
| 1115 |
+
"learning_rate": 1.4346384447674419e-05,
|
| 1116 |
+
"loss": 0.00021764757111668586,
|
| 1117 |
+
"step": 15700
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.4353197674418605,
|
| 1121 |
+
"grad_norm": 0.0011686257785186172,
|
| 1122 |
+
"learning_rate": 1.4119276889534885e-05,
|
| 1123 |
+
"loss": 0.00026009151712059976,
|
| 1124 |
+
"step": 15800
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.4444040697674418,
|
| 1128 |
+
"grad_norm": 0.00034656753996387124,
|
| 1129 |
+
"learning_rate": 1.389216933139535e-05,
|
| 1130 |
+
"loss": 0.0003694407269358635,
|
| 1131 |
+
"step": 15900
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.4534883720930232,
|
| 1135 |
+
"grad_norm": 0.0023575718514621258,
|
| 1136 |
+
"learning_rate": 1.3665061773255814e-05,
|
| 1137 |
+
"loss": 0.0001483263447880745,
|
| 1138 |
+
"step": 16000
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.4625726744186047,
|
| 1142 |
+
"grad_norm": 0.0011886970605701208,
|
| 1143 |
+
"learning_rate": 1.343795421511628e-05,
|
| 1144 |
+
"loss": 0.0004465998336672783,
|
| 1145 |
+
"step": 16100
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.471656976744186,
|
| 1149 |
+
"grad_norm": 0.00039876584196463227,
|
| 1150 |
+
"learning_rate": 1.3210846656976745e-05,
|
| 1151 |
+
"loss": 0.00015990335494279863,
|
| 1152 |
+
"step": 16200
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.4807412790697674,
|
| 1156 |
+
"grad_norm": 0.00023756893642712384,
|
| 1157 |
+
"learning_rate": 1.298373909883721e-05,
|
| 1158 |
+
"loss": 5.0160493701696394e-05,
|
| 1159 |
+
"step": 16300
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.489825581395349,
|
| 1163 |
+
"grad_norm": 0.0002809664292726666,
|
| 1164 |
+
"learning_rate": 1.2756631540697675e-05,
|
| 1165 |
+
"loss": 0.0001377291791141033,
|
| 1166 |
+
"step": 16400
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.4989098837209303,
|
| 1170 |
+
"grad_norm": 0.00037873705150559545,
|
| 1171 |
+
"learning_rate": 1.252952398255814e-05,
|
| 1172 |
+
"loss": 7.196901366114616e-05,
|
| 1173 |
+
"step": 16500
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.5079941860465116,
|
| 1177 |
+
"grad_norm": 1.4199703931808472,
|
| 1178 |
+
"learning_rate": 1.2302416424418606e-05,
|
| 1179 |
+
"loss": 0.0005905186012387276,
|
| 1180 |
+
"step": 16600
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.5170784883720931,
|
| 1184 |
+
"grad_norm": 0.0003422704176045954,
|
| 1185 |
+
"learning_rate": 1.207530886627907e-05,
|
| 1186 |
+
"loss": 0.00013428066857159137,
|
| 1187 |
+
"step": 16700
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.5261627906976745,
|
| 1191 |
+
"grad_norm": 0.0010882618371397257,
|
| 1192 |
+
"learning_rate": 1.1848201308139535e-05,
|
| 1193 |
+
"loss": 0.00020068021491169929,
|
| 1194 |
+
"step": 16800
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.5352470930232558,
|
| 1198 |
+
"grad_norm": 0.024530770257115364,
|
| 1199 |
+
"learning_rate": 1.162109375e-05,
|
| 1200 |
+
"loss": 0.0004120354726910591,
|
| 1201 |
+
"step": 16900
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.5443313953488373,
|
| 1205 |
+
"grad_norm": 0.0006738721276633441,
|
| 1206 |
+
"learning_rate": 1.1393986191860465e-05,
|
| 1207 |
+
"loss": 0.0002767092920839787,
|
| 1208 |
+
"step": 17000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.5534156976744184,
|
| 1212 |
+
"grad_norm": 0.00019564498506952077,
|
| 1213 |
+
"learning_rate": 1.1166878633720931e-05,
|
| 1214 |
+
"loss": 0.00021817052736878396,
|
| 1215 |
+
"step": 17100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.5625,
|
| 1219 |
+
"grad_norm": 0.0002832361788023263,
|
| 1220 |
+
"learning_rate": 1.0939771075581396e-05,
|
| 1221 |
+
"loss": 5.168498028069734e-05,
|
| 1222 |
+
"step": 17200
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.5715843023255816,
|
| 1226 |
+
"grad_norm": 0.0005729420809075236,
|
| 1227 |
+
"learning_rate": 1.071266351744186e-05,
|
| 1228 |
+
"loss": 0.00016681572422385215,
|
| 1229 |
+
"step": 17300
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.5806686046511627,
|
| 1233 |
+
"grad_norm": 0.0007383101619780064,
|
| 1234 |
+
"learning_rate": 1.0485555959302326e-05,
|
| 1235 |
+
"loss": 0.0003441007435321808,
|
| 1236 |
+
"step": 17400
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.5897529069767442,
|
| 1240 |
+
"grad_norm": 0.005897729191929102,
|
| 1241 |
+
"learning_rate": 1.0258448401162791e-05,
|
| 1242 |
+
"loss": 0.0003161391615867615,
|
| 1243 |
+
"step": 17500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.5988372093023255,
|
| 1247 |
+
"grad_norm": 0.00025281202397309244,
|
| 1248 |
+
"learning_rate": 1.0031340843023257e-05,
|
| 1249 |
+
"loss": 0.00011429931037127971,
|
| 1250 |
+
"step": 17600
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.6079215116279069,
|
| 1254 |
+
"grad_norm": 0.0004013874859083444,
|
| 1255 |
+
"learning_rate": 9.804233284883721e-06,
|
| 1256 |
+
"loss": 2.793062012642622e-05,
|
| 1257 |
+
"step": 17700
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.6170058139534884,
|
| 1261 |
+
"grad_norm": 0.008875502273440361,
|
| 1262 |
+
"learning_rate": 9.577125726744186e-06,
|
| 1263 |
+
"loss": 5.882895085960627e-05,
|
| 1264 |
+
"step": 17800
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.6260901162790697,
|
| 1268 |
+
"grad_norm": 0.001823420519940555,
|
| 1269 |
+
"learning_rate": 9.350018168604652e-06,
|
| 1270 |
+
"loss": 0.0002583874017000198,
|
| 1271 |
+
"step": 17900
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.635174418604651,
|
| 1275 |
+
"grad_norm": 0.6227073073387146,
|
| 1276 |
+
"learning_rate": 9.122910610465116e-06,
|
| 1277 |
+
"loss": 0.0007115737348794937,
|
| 1278 |
+
"step": 18000
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.6442587209302326,
|
| 1282 |
+
"grad_norm": 0.0017345056403428316,
|
| 1283 |
+
"learning_rate": 8.895803052325581e-06,
|
| 1284 |
+
"loss": 0.00027986690402030944,
|
| 1285 |
+
"step": 18100
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.653343023255814,
|
| 1289 |
+
"grad_norm": 0.0009033643291331828,
|
| 1290 |
+
"learning_rate": 8.668695494186047e-06,
|
| 1291 |
+
"loss": 0.0002979452162981033,
|
| 1292 |
+
"step": 18200
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.6624273255813953,
|
| 1296 |
+
"grad_norm": 0.00028923453646712005,
|
| 1297 |
+
"learning_rate": 8.441587936046511e-06,
|
| 1298 |
+
"loss": 1.926603843457997e-05,
|
| 1299 |
+
"step": 18300
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.6715116279069768,
|
| 1303 |
+
"grad_norm": 0.0006103311898186803,
|
| 1304 |
+
"learning_rate": 8.214480377906978e-06,
|
| 1305 |
+
"loss": 6.331724114716053e-05,
|
| 1306 |
+
"step": 18400
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.6805959302325582,
|
| 1310 |
+
"grad_norm": 0.0002818437642417848,
|
| 1311 |
+
"learning_rate": 7.987372819767442e-06,
|
| 1312 |
+
"loss": 0.00014105773530900478,
|
| 1313 |
+
"step": 18500
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.6896802325581395,
|
| 1317 |
+
"grad_norm": 0.0011020454112440348,
|
| 1318 |
+
"learning_rate": 7.760265261627907e-06,
|
| 1319 |
+
"loss": 0.0002292773686349392,
|
| 1320 |
+
"step": 18600
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.698764534883721,
|
| 1324 |
+
"grad_norm": 0.001500884653069079,
|
| 1325 |
+
"learning_rate": 7.533157703488372e-06,
|
| 1326 |
+
"loss": 0.00011514685116708279,
|
| 1327 |
+
"step": 18700
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1.7078488372093024,
|
| 1331 |
+
"grad_norm": 0.00016000888717826456,
|
| 1332 |
+
"learning_rate": 7.306050145348838e-06,
|
| 1333 |
+
"loss": 2.0523781422525645e-05,
|
| 1334 |
+
"step": 18800
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.7169331395348837,
|
| 1338 |
+
"grad_norm": 0.003748674876987934,
|
| 1339 |
+
"learning_rate": 7.078942587209303e-06,
|
| 1340 |
+
"loss": 0.00043839264661073686,
|
| 1341 |
+
"step": 18900
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.7260174418604652,
|
| 1345 |
+
"grad_norm": 0.06772974133491516,
|
| 1346 |
+
"learning_rate": 6.8518350290697685e-06,
|
| 1347 |
+
"loss": 2.126413397490978e-05,
|
| 1348 |
+
"step": 19000
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.7351017441860463,
|
| 1352 |
+
"grad_norm": 0.0004789210797753185,
|
| 1353 |
+
"learning_rate": 6.624727470930232e-06,
|
| 1354 |
+
"loss": 0.00020929597318172455,
|
| 1355 |
+
"step": 19100
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.744186046511628,
|
| 1359 |
+
"grad_norm": 7.055519381538033e-05,
|
| 1360 |
+
"learning_rate": 6.397619912790697e-06,
|
| 1361 |
+
"loss": 0.00032072752714157104,
|
| 1362 |
+
"step": 19200
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.7532703488372094,
|
| 1366 |
+
"grad_norm": 0.00010641128028510138,
|
| 1367 |
+
"learning_rate": 6.1705123546511635e-06,
|
| 1368 |
+
"loss": 8.132393471896649e-05,
|
| 1369 |
+
"step": 19300
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.7623546511627906,
|
| 1373 |
+
"grad_norm": 0.0005615473492071033,
|
| 1374 |
+
"learning_rate": 5.943404796511629e-06,
|
| 1375 |
+
"loss": 9.162256610579788e-06,
|
| 1376 |
+
"step": 19400
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.771438953488372,
|
| 1380 |
+
"grad_norm": 0.004505404736846685,
|
| 1381 |
+
"learning_rate": 5.716297238372093e-06,
|
| 1382 |
+
"loss": 9.514865465462207e-05,
|
| 1383 |
+
"step": 19500
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.7805232558139537,
|
| 1387 |
+
"grad_norm": 0.0011890050955116749,
|
| 1388 |
+
"learning_rate": 5.4891896802325586e-06,
|
| 1389 |
+
"loss": 3.7443286273628474e-05,
|
| 1390 |
+
"step": 19600
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.7896075581395348,
|
| 1394 |
+
"grad_norm": 0.000441042153397575,
|
| 1395 |
+
"learning_rate": 5.262082122093023e-06,
|
| 1396 |
+
"loss": 5.087008234113455e-05,
|
| 1397 |
+
"step": 19700
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.7986918604651163,
|
| 1401 |
+
"grad_norm": 0.0001777316356310621,
|
| 1402 |
+
"learning_rate": 5.034974563953489e-06,
|
| 1403 |
+
"loss": 0.00017192648723721505,
|
| 1404 |
+
"step": 19800
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.8077761627906976,
|
| 1408 |
+
"grad_norm": 0.00015439293929375708,
|
| 1409 |
+
"learning_rate": 4.8078670058139536e-06,
|
| 1410 |
+
"loss": 0.0001058769691735506,
|
| 1411 |
+
"step": 19900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 1.816860465116279,
|
| 1415 |
+
"grad_norm": 0.0016349812503904104,
|
| 1416 |
+
"learning_rate": 4.580759447674419e-06,
|
| 1417 |
+
"loss": 1.0620863176882267e-05,
|
| 1418 |
+
"step": 20000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.8259447674418605,
|
| 1422 |
+
"grad_norm": 9.556530858390033e-05,
|
| 1423 |
+
"learning_rate": 4.353651889534884e-06,
|
| 1424 |
+
"loss": 0.00019107908010482789,
|
| 1425 |
+
"step": 20100
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.8350290697674418,
|
| 1429 |
+
"grad_norm": 0.0002568990457803011,
|
| 1430 |
+
"learning_rate": 4.126544331395349e-06,
|
| 1431 |
+
"loss": 0.00026563439518213273,
|
| 1432 |
+
"step": 20200
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.8441133720930232,
|
| 1436 |
+
"grad_norm": 0.5719628930091858,
|
| 1437 |
+
"learning_rate": 3.899436773255814e-06,
|
| 1438 |
+
"loss": 5.403281655162573e-05,
|
| 1439 |
+
"step": 20300
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.8531976744186047,
|
| 1443 |
+
"grad_norm": 0.017383404076099396,
|
| 1444 |
+
"learning_rate": 3.672329215116279e-06,
|
| 1445 |
+
"loss": 0.00036652404814958573,
|
| 1446 |
+
"step": 20400
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.862281976744186,
|
| 1450 |
+
"grad_norm": 0.00022910887491889298,
|
| 1451 |
+
"learning_rate": 3.4452216569767445e-06,
|
| 1452 |
+
"loss": 0.000403200164437294,
|
| 1453 |
+
"step": 20500
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.8713662790697674,
|
| 1457 |
+
"grad_norm": 0.00033295468892902136,
|
| 1458 |
+
"learning_rate": 3.2181140988372097e-06,
|
| 1459 |
+
"loss": 8.646129630506039e-05,
|
| 1460 |
+
"step": 20600
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.880450581395349,
|
| 1464 |
+
"grad_norm": 0.00012523184705059975,
|
| 1465 |
+
"learning_rate": 2.9910065406976746e-06,
|
| 1466 |
+
"loss": 0.00016864996403455733,
|
| 1467 |
+
"step": 20700
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.8895348837209303,
|
| 1471 |
+
"grad_norm": 0.00022476979938801378,
|
| 1472 |
+
"learning_rate": 2.7638989825581395e-06,
|
| 1473 |
+
"loss": 5.167209077626467e-05,
|
| 1474 |
+
"step": 20800
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.8986191860465116,
|
| 1478 |
+
"grad_norm": 0.0002389108412899077,
|
| 1479 |
+
"learning_rate": 2.5367914244186048e-06,
|
| 1480 |
+
"loss": 6.591790355741977e-05,
|
| 1481 |
+
"step": 20900
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.9077034883720931,
|
| 1485 |
+
"grad_norm": 0.0013002109481021762,
|
| 1486 |
+
"learning_rate": 2.30968386627907e-06,
|
| 1487 |
+
"loss": 0.00010993644595146179,
|
| 1488 |
+
"step": 21000
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.9167877906976745,
|
| 1492 |
+
"grad_norm": 0.13188259303569794,
|
| 1493 |
+
"learning_rate": 2.082576308139535e-06,
|
| 1494 |
+
"loss": 0.00014082306995987892,
|
| 1495 |
+
"step": 21100
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.9258720930232558,
|
| 1499 |
+
"grad_norm": 0.0010737127158790827,
|
| 1500 |
+
"learning_rate": 1.85546875e-06,
|
| 1501 |
+
"loss": 0.00030819986015558244,
|
| 1502 |
+
"step": 21200
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.9349563953488373,
|
| 1506 |
+
"grad_norm": 7.047707185847685e-05,
|
| 1507 |
+
"learning_rate": 1.628361191860465e-06,
|
| 1508 |
+
"loss": 2.0567586179822683e-05,
|
| 1509 |
+
"step": 21300
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.9440406976744184,
|
| 1513 |
+
"grad_norm": 0.0001350079692201689,
|
| 1514 |
+
"learning_rate": 1.4012536337209304e-06,
|
| 1515 |
+
"loss": 7.989832083694636e-06,
|
| 1516 |
+
"step": 21400
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.953125,
|
| 1520 |
+
"grad_norm": 0.0008243785705417395,
|
| 1521 |
+
"learning_rate": 1.1741460755813954e-06,
|
| 1522 |
+
"loss": 0.0002570002153515816,
|
| 1523 |
+
"step": 21500
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.9622093023255816,
|
| 1527 |
+
"grad_norm": 0.001545518171042204,
|
| 1528 |
+
"learning_rate": 9.470385174418604e-07,
|
| 1529 |
+
"loss": 2.7221005875617266e-05,
|
| 1530 |
+
"step": 21600
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 1.9712936046511627,
|
| 1534 |
+
"grad_norm": 5.5686323321424425e-05,
|
| 1535 |
+
"learning_rate": 7.199309593023256e-07,
|
| 1536 |
+
"loss": 0.00010406752116978168,
|
| 1537 |
+
"step": 21700
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 1.9803779069767442,
|
| 1541 |
+
"grad_norm": 5.525942106032744e-05,
|
| 1542 |
+
"learning_rate": 4.928234011627908e-07,
|
| 1543 |
+
"loss": 3.853037022054195e-05,
|
| 1544 |
+
"step": 21800
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.9894622093023255,
|
| 1548 |
+
"grad_norm": 0.0002285480877617374,
|
| 1549 |
+
"learning_rate": 2.657158430232558e-07,
|
| 1550 |
+
"loss": 0.00048015639185905457,
|
| 1551 |
+
"step": 21900
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.9985465116279069,
|
| 1555 |
+
"grad_norm": 8.369733404833823e-05,
|
| 1556 |
+
"learning_rate": 3.860828488372093e-08,
|
| 1557 |
+
"loss": 6.105023785494268e-06,
|
| 1558 |
+
"step": 22000
|
| 1559 |
+
}
|
| 1560 |
+
],
|
| 1561 |
+
"logging_steps": 100,
|
| 1562 |
+
"max_steps": 22016,
|
| 1563 |
+
"num_input_tokens_seen": 0,
|
| 1564 |
+
"num_train_epochs": 2,
|
| 1565 |
+
"save_steps": 500,
|
| 1566 |
+
"stateful_callbacks": {
|
| 1567 |
+
"TrainerControl": {
|
| 1568 |
+
"args": {
|
| 1569 |
+
"should_epoch_stop": false,
|
| 1570 |
+
"should_evaluate": false,
|
| 1571 |
+
"should_log": false,
|
| 1572 |
+
"should_save": true,
|
| 1573 |
+
"should_training_stop": true
|
| 1574 |
+
},
|
| 1575 |
+
"attributes": {}
|
| 1576 |
+
}
|
| 1577 |
+
},
|
| 1578 |
+
"total_flos": 1.342304325992448e+16,
|
| 1579 |
+
"train_batch_size": 8,
|
| 1580 |
+
"trial_name": null,
|
| 1581 |
+
"trial_params": null
|
| 1582 |
+
}
|
models/dostoievsky_v1/checkpoint-22016/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:edab2539a81b65c2a76c4c39bed219aa02358f44f3286429f83165442a1b53fa
|
| 3 |
+
size 5329
|
models/dostoievsky_v1/final_model/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.1,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"add_bias_logits": false,
|
| 5 |
+
"add_final_layer_norm": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"BartForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.1,
|
| 10 |
+
"bos_token_id": 0,
|
| 11 |
+
"classif_dropout": 0.1,
|
| 12 |
+
"classifier_dropout": 0.0,
|
| 13 |
+
"d_model": 768,
|
| 14 |
+
"decoder_attention_heads": 12,
|
| 15 |
+
"decoder_ffn_dim": 3072,
|
| 16 |
+
"decoder_layerdrop": 0.0,
|
| 17 |
+
"decoder_layers": 6,
|
| 18 |
+
"decoder_start_token_id": 2,
|
| 19 |
+
"dropout": 0.1,
|
| 20 |
+
"dtype": "float32",
|
| 21 |
+
"encoder_attention_heads": 12,
|
| 22 |
+
"encoder_ffn_dim": 3072,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 6,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": true,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1,
|
| 39 |
+
"LABEL_2": 2
|
| 40 |
+
},
|
| 41 |
+
"max_position_embeddings": 1024,
|
| 42 |
+
"model_type": "bart",
|
| 43 |
+
"normalize_before": false,
|
| 44 |
+
"normalize_embedding": true,
|
| 45 |
+
"pad_token_id": 1,
|
| 46 |
+
"scale_embedding": false,
|
| 47 |
+
"task_specific_params": {
|
| 48 |
+
"summarization": {
|
| 49 |
+
"length_penalty": 1.0,
|
| 50 |
+
"max_length": 128,
|
| 51 |
+
"min_length": 12,
|
| 52 |
+
"num_beams": 4
|
| 53 |
+
},
|
| 54 |
+
"summarization_cnn": {
|
| 55 |
+
"length_penalty": 2.0,
|
| 56 |
+
"max_length": 142,
|
| 57 |
+
"min_length": 56,
|
| 58 |
+
"num_beams": 4
|
| 59 |
+
},
|
| 60 |
+
"summarization_xsum": {
|
| 61 |
+
"length_penalty": 1.0,
|
| 62 |
+
"max_length": 62,
|
| 63 |
+
"min_length": 11,
|
| 64 |
+
"num_beams": 6
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.5.0",
|
| 69 |
+
"use_cache": false,
|
| 70 |
+
"vocab_size": 50265
|
| 71 |
+
}
|
models/dostoievsky_v1/final_model/generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"early_stopping": true,
|
| 6 |
+
"eos_token_id": [
|
| 7 |
+
2
|
| 8 |
+
],
|
| 9 |
+
"forced_bos_token_id": 0,
|
| 10 |
+
"forced_eos_token_id": 2,
|
| 11 |
+
"no_repeat_ngram_size": 3,
|
| 12 |
+
"num_beams": 4,
|
| 13 |
+
"pad_token_id": 1,
|
| 14 |
+
"transformers_version": "5.5.0",
|
| 15 |
+
"use_cache": true
|
| 16 |
+
}
|
models/dostoievsky_v1/final_model/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d9dc9ffae85920773b2b5774f37b9630b197fc455008db011e6288a451870cb8
|
| 3 |
+
size 557912620
|
models/dostoievsky_v1/final_model/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
models/dostoievsky_v1/final_model/tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"mask_token": "<mask>",
|
| 10 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 11 |
+
"pad_token": "<pad>",
|
| 12 |
+
"sep_token": "</s>",
|
| 13 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 14 |
+
"trim_offsets": true,
|
| 15 |
+
"unk_token": "<unk>"
|
| 16 |
+
}
|
models/dostoievsky_v1/final_model/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:edab2539a81b65c2a76c4c39bed219aa02358f44f3286429f83165442a1b53fa
|
| 3 |
+
size 5329
|