Commit
•
0a18762
1
Parent(s):
a0acee2
Update version 3 of T5 (#25)
Browse files- Update version 3 of T5 (0d1e26f60dfdb3c4f4a96eb049022798f3ae5bd3)
Co-authored-by: Hoang Trung Khai <[email protected]>
- T5_ver3/config.json +61 -0
- T5_ver3/generation_config.json +7 -0
- T5_ver3/model.safetensors +3 -0
- T5_ver3/optimizer.pt +3 -0
- T5_ver3/rng_state.pth +3 -0
- T5_ver3/scheduler.pt +3 -0
- T5_ver3/trainer_state.json +956 -0
- T5_ver3/training_args.bin +3 -0
- T5_ver3/vi_tokenizer_32128.model +3 -0
T5_ver3/config.json
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "t5-small",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"T5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"classifier_dropout": 0.0,
|
| 7 |
+
"d_ff": 2048,
|
| 8 |
+
"d_kv": 64,
|
| 9 |
+
"d_model": 512,
|
| 10 |
+
"decoder_start_token_id": 0,
|
| 11 |
+
"dense_act_fn": "relu",
|
| 12 |
+
"dropout_rate": 0.1,
|
| 13 |
+
"eos_token_id": 1,
|
| 14 |
+
"feed_forward_proj": "relu",
|
| 15 |
+
"initializer_factor": 1.0,
|
| 16 |
+
"is_encoder_decoder": true,
|
| 17 |
+
"is_gated_act": false,
|
| 18 |
+
"layer_norm_epsilon": 1e-06,
|
| 19 |
+
"model_type": "t5",
|
| 20 |
+
"n_positions": 512,
|
| 21 |
+
"num_decoder_layers": 6,
|
| 22 |
+
"num_heads": 8,
|
| 23 |
+
"num_layers": 6,
|
| 24 |
+
"output_past": true,
|
| 25 |
+
"pad_token_id": 0,
|
| 26 |
+
"relative_attention_max_distance": 128,
|
| 27 |
+
"relative_attention_num_buckets": 32,
|
| 28 |
+
"task_specific_params": {
|
| 29 |
+
"summarization": {
|
| 30 |
+
"early_stopping": true,
|
| 31 |
+
"length_penalty": 2.0,
|
| 32 |
+
"max_length": 200,
|
| 33 |
+
"min_length": 30,
|
| 34 |
+
"no_repeat_ngram_size": 3,
|
| 35 |
+
"num_beams": 4,
|
| 36 |
+
"prefix": "summarize: "
|
| 37 |
+
},
|
| 38 |
+
"translation_en_to_de": {
|
| 39 |
+
"early_stopping": true,
|
| 40 |
+
"max_length": 300,
|
| 41 |
+
"num_beams": 4,
|
| 42 |
+
"prefix": "translate English to German: "
|
| 43 |
+
},
|
| 44 |
+
"translation_en_to_fr": {
|
| 45 |
+
"early_stopping": true,
|
| 46 |
+
"max_length": 300,
|
| 47 |
+
"num_beams": 4,
|
| 48 |
+
"prefix": "translate English to French: "
|
| 49 |
+
},
|
| 50 |
+
"translation_en_to_ro": {
|
| 51 |
+
"early_stopping": true,
|
| 52 |
+
"max_length": 300,
|
| 53 |
+
"num_beams": 4,
|
| 54 |
+
"prefix": "translate English to Romanian: "
|
| 55 |
+
}
|
| 56 |
+
},
|
| 57 |
+
"torch_dtype": "float32",
|
| 58 |
+
"transformers_version": "4.46.3",
|
| 59 |
+
"use_cache": true,
|
| 60 |
+
"vocab_size": 32128
|
| 61 |
+
}
|
T5_ver3/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.46.3"
|
| 7 |
+
}
|
T5_ver3/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a99d19bb740d05d7ba6fe9fc6a29c1fb87b31c5f53abc8069481b676486b9f6
|
| 3 |
+
size 242041896
|
T5_ver3/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d621ccaea474b7af718e689d55c3a2f8b3eaceb044beba2a95d2431f4a32e85b
|
| 3 |
+
size 484163514
|
T5_ver3/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62a5c588d4a4472691da0055c5fbfc23292d6d762491ad1788c43575772f8c23
|
| 3 |
+
size 14244
|
T5_ver3/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7322999c5b4914592281af4aa5c83121ae85b9aa78eea758a9b372cfcd470c0a
|
| 3 |
+
size 1064
|
T5_ver3/trainer_state.json
ADDED
|
@@ -0,0 +1,956 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8676381707191467,
|
| 3 |
+
"best_model_checkpoint": "./results_v1/checkpoint-10200",
|
| 4 |
+
"epoch": 25.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.24509803921568626,
|
| 13 |
+
"grad_norm": 10681.4521484375,
|
| 14 |
+
"learning_rate": 0.0004950980392156863,
|
| 15 |
+
"loss": 2.5241,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.49019607843137253,
|
| 20 |
+
"grad_norm": 6339.50927734375,
|
| 21 |
+
"learning_rate": 0.0004901960784313725,
|
| 22 |
+
"loss": 2.0561,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.7352941176470589,
|
| 27 |
+
"grad_norm": 16086.7705078125,
|
| 28 |
+
"learning_rate": 0.0004852941176470588,
|
| 29 |
+
"loss": 1.9379,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.9803921568627451,
|
| 34 |
+
"grad_norm": 10268.5185546875,
|
| 35 |
+
"learning_rate": 0.00048039215686274513,
|
| 36 |
+
"loss": 1.8333,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.0,
|
| 41 |
+
"eval_loss": 1.6950186491012573,
|
| 42 |
+
"eval_runtime": 17.2704,
|
| 43 |
+
"eval_samples_per_second": 463.222,
|
| 44 |
+
"eval_steps_per_second": 3.648,
|
| 45 |
+
"step": 408
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.2254901960784315,
|
| 49 |
+
"grad_norm": 10859.7197265625,
|
| 50 |
+
"learning_rate": 0.00047549019607843134,
|
| 51 |
+
"loss": 1.7372,
|
| 52 |
+
"step": 500
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.4705882352941178,
|
| 56 |
+
"grad_norm": 13147.400390625,
|
| 57 |
+
"learning_rate": 0.00047058823529411766,
|
| 58 |
+
"loss": 1.6714,
|
| 59 |
+
"step": 600
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.715686274509804,
|
| 63 |
+
"grad_norm": 11227.5166015625,
|
| 64 |
+
"learning_rate": 0.0004656862745098039,
|
| 65 |
+
"loss": 1.5845,
|
| 66 |
+
"step": 700
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.9607843137254903,
|
| 70 |
+
"grad_norm": 10829.8447265625,
|
| 71 |
+
"learning_rate": 0.0004607843137254902,
|
| 72 |
+
"loss": 1.5283,
|
| 73 |
+
"step": 800
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 2.0,
|
| 77 |
+
"eval_loss": 1.389205813407898,
|
| 78 |
+
"eval_runtime": 16.6762,
|
| 79 |
+
"eval_samples_per_second": 479.726,
|
| 80 |
+
"eval_steps_per_second": 3.778,
|
| 81 |
+
"step": 816
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.2058823529411766,
|
| 85 |
+
"grad_norm": 13556.080078125,
|
| 86 |
+
"learning_rate": 0.00045588235294117646,
|
| 87 |
+
"loss": 1.4638,
|
| 88 |
+
"step": 900
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 2.450980392156863,
|
| 92 |
+
"grad_norm": 10844.283203125,
|
| 93 |
+
"learning_rate": 0.0004509803921568628,
|
| 94 |
+
"loss": 1.4344,
|
| 95 |
+
"step": 1000
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.696078431372549,
|
| 99 |
+
"grad_norm": 12495.66796875,
|
| 100 |
+
"learning_rate": 0.00044607843137254904,
|
| 101 |
+
"loss": 1.3873,
|
| 102 |
+
"step": 1100
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.9411764705882355,
|
| 106 |
+
"grad_norm": 11954.7265625,
|
| 107 |
+
"learning_rate": 0.0004411764705882353,
|
| 108 |
+
"loss": 1.3733,
|
| 109 |
+
"step": 1200
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 3.0,
|
| 113 |
+
"eval_loss": 1.241082787513733,
|
| 114 |
+
"eval_runtime": 17.1795,
|
| 115 |
+
"eval_samples_per_second": 465.671,
|
| 116 |
+
"eval_steps_per_second": 3.667,
|
| 117 |
+
"step": 1224
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 3.186274509803922,
|
| 121 |
+
"grad_norm": 12640.005859375,
|
| 122 |
+
"learning_rate": 0.0004362745098039216,
|
| 123 |
+
"loss": 1.32,
|
| 124 |
+
"step": 1300
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 3.431372549019608,
|
| 128 |
+
"grad_norm": 14839.830078125,
|
| 129 |
+
"learning_rate": 0.0004313725490196079,
|
| 130 |
+
"loss": 1.2993,
|
| 131 |
+
"step": 1400
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.6764705882352944,
|
| 135 |
+
"grad_norm": 17634.357421875,
|
| 136 |
+
"learning_rate": 0.0004264705882352941,
|
| 137 |
+
"loss": 1.284,
|
| 138 |
+
"step": 1500
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.9215686274509802,
|
| 142 |
+
"grad_norm": 11596.447265625,
|
| 143 |
+
"learning_rate": 0.00042156862745098043,
|
| 144 |
+
"loss": 1.2622,
|
| 145 |
+
"step": 1600
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 4.0,
|
| 149 |
+
"eval_loss": 1.152040958404541,
|
| 150 |
+
"eval_runtime": 16.7185,
|
| 151 |
+
"eval_samples_per_second": 478.513,
|
| 152 |
+
"eval_steps_per_second": 3.768,
|
| 153 |
+
"step": 1632
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 4.166666666666667,
|
| 157 |
+
"grad_norm": 10679.4697265625,
|
| 158 |
+
"learning_rate": 0.0004166666666666667,
|
| 159 |
+
"loss": 1.2385,
|
| 160 |
+
"step": 1700
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 4.411764705882353,
|
| 164 |
+
"grad_norm": 10487.45703125,
|
| 165 |
+
"learning_rate": 0.0004117647058823529,
|
| 166 |
+
"loss": 1.2131,
|
| 167 |
+
"step": 1800
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 4.6568627450980395,
|
| 171 |
+
"grad_norm": 16033.8701171875,
|
| 172 |
+
"learning_rate": 0.0004068627450980392,
|
| 173 |
+
"loss": 1.2052,
|
| 174 |
+
"step": 1900
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 4.901960784313726,
|
| 178 |
+
"grad_norm": 13594.8916015625,
|
| 179 |
+
"learning_rate": 0.0004019607843137255,
|
| 180 |
+
"loss": 1.1878,
|
| 181 |
+
"step": 2000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 5.0,
|
| 185 |
+
"eval_loss": 1.0944446325302124,
|
| 186 |
+
"eval_runtime": 16.7053,
|
| 187 |
+
"eval_samples_per_second": 478.89,
|
| 188 |
+
"eval_steps_per_second": 3.771,
|
| 189 |
+
"step": 2040
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 5.147058823529412,
|
| 193 |
+
"grad_norm": 12540.892578125,
|
| 194 |
+
"learning_rate": 0.00039705882352941176,
|
| 195 |
+
"loss": 1.1714,
|
| 196 |
+
"step": 2100
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 5.392156862745098,
|
| 200 |
+
"grad_norm": 11532.8291015625,
|
| 201 |
+
"learning_rate": 0.000392156862745098,
|
| 202 |
+
"loss": 1.1554,
|
| 203 |
+
"step": 2200
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 5.637254901960784,
|
| 207 |
+
"grad_norm": 13462.1943359375,
|
| 208 |
+
"learning_rate": 0.00038725490196078434,
|
| 209 |
+
"loss": 1.1487,
|
| 210 |
+
"step": 2300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 5.882352941176471,
|
| 214 |
+
"grad_norm": 14919.408203125,
|
| 215 |
+
"learning_rate": 0.00038235294117647055,
|
| 216 |
+
"loss": 1.1286,
|
| 217 |
+
"step": 2400
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 6.0,
|
| 221 |
+
"eval_loss": 1.0514171123504639,
|
| 222 |
+
"eval_runtime": 17.2713,
|
| 223 |
+
"eval_samples_per_second": 463.195,
|
| 224 |
+
"eval_steps_per_second": 3.648,
|
| 225 |
+
"step": 2448
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 6.127450980392156,
|
| 229 |
+
"grad_norm": 12428.1318359375,
|
| 230 |
+
"learning_rate": 0.0003774509803921569,
|
| 231 |
+
"loss": 1.1157,
|
| 232 |
+
"step": 2500
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 6.372549019607844,
|
| 236 |
+
"grad_norm": 12143.3505859375,
|
| 237 |
+
"learning_rate": 0.00037254901960784314,
|
| 238 |
+
"loss": 1.1092,
|
| 239 |
+
"step": 2600
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 6.617647058823529,
|
| 243 |
+
"grad_norm": 11514.5810546875,
|
| 244 |
+
"learning_rate": 0.00036764705882352946,
|
| 245 |
+
"loss": 1.0906,
|
| 246 |
+
"step": 2700
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 6.862745098039216,
|
| 250 |
+
"grad_norm": 18561.646484375,
|
| 251 |
+
"learning_rate": 0.00036274509803921567,
|
| 252 |
+
"loss": 1.0951,
|
| 253 |
+
"step": 2800
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 7.0,
|
| 257 |
+
"eval_loss": 1.0217130184173584,
|
| 258 |
+
"eval_runtime": 16.6786,
|
| 259 |
+
"eval_samples_per_second": 479.657,
|
| 260 |
+
"eval_steps_per_second": 3.777,
|
| 261 |
+
"step": 2856
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 7.107843137254902,
|
| 265 |
+
"grad_norm": 16182.7216796875,
|
| 266 |
+
"learning_rate": 0.000357843137254902,
|
| 267 |
+
"loss": 1.0806,
|
| 268 |
+
"step": 2900
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 7.352941176470588,
|
| 272 |
+
"grad_norm": 15809.748046875,
|
| 273 |
+
"learning_rate": 0.00035294117647058826,
|
| 274 |
+
"loss": 1.0551,
|
| 275 |
+
"step": 3000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 7.598039215686274,
|
| 279 |
+
"grad_norm": 11234.1923828125,
|
| 280 |
+
"learning_rate": 0.0003480392156862745,
|
| 281 |
+
"loss": 1.0648,
|
| 282 |
+
"step": 3100
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 7.8431372549019605,
|
| 286 |
+
"grad_norm": 13925.9716796875,
|
| 287 |
+
"learning_rate": 0.0003431372549019608,
|
| 288 |
+
"loss": 1.0551,
|
| 289 |
+
"step": 3200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 8.0,
|
| 293 |
+
"eval_loss": 0.9909818172454834,
|
| 294 |
+
"eval_runtime": 16.7369,
|
| 295 |
+
"eval_samples_per_second": 477.987,
|
| 296 |
+
"eval_steps_per_second": 3.764,
|
| 297 |
+
"step": 3264
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 8.088235294117647,
|
| 301 |
+
"grad_norm": 13518.5712890625,
|
| 302 |
+
"learning_rate": 0.0003382352941176471,
|
| 303 |
+
"loss": 1.0471,
|
| 304 |
+
"step": 3300
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 8.333333333333334,
|
| 308 |
+
"grad_norm": 12719.046875,
|
| 309 |
+
"learning_rate": 0.0003333333333333333,
|
| 310 |
+
"loss": 1.0271,
|
| 311 |
+
"step": 3400
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 8.57843137254902,
|
| 315 |
+
"grad_norm": 13152.646484375,
|
| 316 |
+
"learning_rate": 0.0003284313725490196,
|
| 317 |
+
"loss": 1.0161,
|
| 318 |
+
"step": 3500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 8.823529411764707,
|
| 322 |
+
"grad_norm": 11174.1962890625,
|
| 323 |
+
"learning_rate": 0.0003235294117647059,
|
| 324 |
+
"loss": 1.0259,
|
| 325 |
+
"step": 3600
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 9.0,
|
| 329 |
+
"eval_loss": 0.9708221554756165,
|
| 330 |
+
"eval_runtime": 17.2479,
|
| 331 |
+
"eval_samples_per_second": 463.825,
|
| 332 |
+
"eval_steps_per_second": 3.653,
|
| 333 |
+
"step": 3672
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 9.068627450980392,
|
| 337 |
+
"grad_norm": 12128.1474609375,
|
| 338 |
+
"learning_rate": 0.0003186274509803921,
|
| 339 |
+
"loss": 1.0239,
|
| 340 |
+
"step": 3700
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 9.313725490196079,
|
| 344 |
+
"grad_norm": 10996.591796875,
|
| 345 |
+
"learning_rate": 0.00031372549019607844,
|
| 346 |
+
"loss": 0.9963,
|
| 347 |
+
"step": 3800
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 9.558823529411764,
|
| 351 |
+
"grad_norm": 12174.8466796875,
|
| 352 |
+
"learning_rate": 0.0003088235294117647,
|
| 353 |
+
"loss": 1.006,
|
| 354 |
+
"step": 3900
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 9.803921568627452,
|
| 358 |
+
"grad_norm": 11552.1796875,
|
| 359 |
+
"learning_rate": 0.00030392156862745097,
|
| 360 |
+
"loss": 1.0001,
|
| 361 |
+
"step": 4000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 10.0,
|
| 365 |
+
"eval_loss": 0.9560179114341736,
|
| 366 |
+
"eval_runtime": 17.2671,
|
| 367 |
+
"eval_samples_per_second": 463.308,
|
| 368 |
+
"eval_steps_per_second": 3.649,
|
| 369 |
+
"step": 4080
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 10.049019607843137,
|
| 373 |
+
"grad_norm": 14723.912109375,
|
| 374 |
+
"learning_rate": 0.00029901960784313724,
|
| 375 |
+
"loss": 0.9901,
|
| 376 |
+
"step": 4100
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 10.294117647058824,
|
| 380 |
+
"grad_norm": 16030.7060546875,
|
| 381 |
+
"learning_rate": 0.00029411764705882356,
|
| 382 |
+
"loss": 0.9763,
|
| 383 |
+
"step": 4200
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 10.53921568627451,
|
| 387 |
+
"grad_norm": 12146.142578125,
|
| 388 |
+
"learning_rate": 0.0002892156862745098,
|
| 389 |
+
"loss": 0.9638,
|
| 390 |
+
"step": 4300
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 10.784313725490197,
|
| 394 |
+
"grad_norm": 12161.0078125,
|
| 395 |
+
"learning_rate": 0.0002843137254901961,
|
| 396 |
+
"loss": 0.9891,
|
| 397 |
+
"step": 4400
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 11.0,
|
| 401 |
+
"eval_loss": 0.9404354095458984,
|
| 402 |
+
"eval_runtime": 16.7426,
|
| 403 |
+
"eval_samples_per_second": 477.824,
|
| 404 |
+
"eval_steps_per_second": 3.763,
|
| 405 |
+
"step": 4488
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 11.029411764705882,
|
| 409 |
+
"grad_norm": 11092.498046875,
|
| 410 |
+
"learning_rate": 0.00027941176470588236,
|
| 411 |
+
"loss": 0.9653,
|
| 412 |
+
"step": 4500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 11.27450980392157,
|
| 416 |
+
"grad_norm": 15073.44140625,
|
| 417 |
+
"learning_rate": 0.0002745098039215687,
|
| 418 |
+
"loss": 0.9608,
|
| 419 |
+
"step": 4600
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 11.519607843137255,
|
| 423 |
+
"grad_norm": 12903.4189453125,
|
| 424 |
+
"learning_rate": 0.0002696078431372549,
|
| 425 |
+
"loss": 0.9572,
|
| 426 |
+
"step": 4700
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 11.764705882352942,
|
| 430 |
+
"grad_norm": 14061.0234375,
|
| 431 |
+
"learning_rate": 0.0002647058823529412,
|
| 432 |
+
"loss": 0.9523,
|
| 433 |
+
"step": 4800
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 12.0,
|
| 437 |
+
"eval_loss": 0.9268321394920349,
|
| 438 |
+
"eval_runtime": 16.7172,
|
| 439 |
+
"eval_samples_per_second": 478.548,
|
| 440 |
+
"eval_steps_per_second": 3.769,
|
| 441 |
+
"step": 4896
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 12.009803921568627,
|
| 445 |
+
"grad_norm": 13276.89453125,
|
| 446 |
+
"learning_rate": 0.0002598039215686275,
|
| 447 |
+
"loss": 0.9492,
|
| 448 |
+
"step": 4900
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 12.254901960784313,
|
| 452 |
+
"grad_norm": 11237.9267578125,
|
| 453 |
+
"learning_rate": 0.00025490196078431374,
|
| 454 |
+
"loss": 0.9417,
|
| 455 |
+
"step": 5000
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 12.5,
|
| 459 |
+
"grad_norm": 11576.23046875,
|
| 460 |
+
"learning_rate": 0.00025,
|
| 461 |
+
"loss": 0.9379,
|
| 462 |
+
"step": 5100
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 12.745098039215687,
|
| 466 |
+
"grad_norm": 14009.544921875,
|
| 467 |
+
"learning_rate": 0.00024509803921568627,
|
| 468 |
+
"loss": 0.9349,
|
| 469 |
+
"step": 5200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 12.990196078431373,
|
| 473 |
+
"grad_norm": 12153.9677734375,
|
| 474 |
+
"learning_rate": 0.00024019607843137256,
|
| 475 |
+
"loss": 0.9329,
|
| 476 |
+
"step": 5300
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 13.0,
|
| 480 |
+
"eval_loss": 0.9184120893478394,
|
| 481 |
+
"eval_runtime": 17.2276,
|
| 482 |
+
"eval_samples_per_second": 464.372,
|
| 483 |
+
"eval_steps_per_second": 3.657,
|
| 484 |
+
"step": 5304
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 13.235294117647058,
|
| 488 |
+
"grad_norm": 12770.306640625,
|
| 489 |
+
"learning_rate": 0.00023529411764705883,
|
| 490 |
+
"loss": 0.9219,
|
| 491 |
+
"step": 5400
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 13.480392156862745,
|
| 495 |
+
"grad_norm": 12070.76171875,
|
| 496 |
+
"learning_rate": 0.0002303921568627451,
|
| 497 |
+
"loss": 0.9192,
|
| 498 |
+
"step": 5500
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 13.72549019607843,
|
| 502 |
+
"grad_norm": 12918.3779296875,
|
| 503 |
+
"learning_rate": 0.0002254901960784314,
|
| 504 |
+
"loss": 0.925,
|
| 505 |
+
"step": 5600
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 13.970588235294118,
|
| 509 |
+
"grad_norm": 11361.103515625,
|
| 510 |
+
"learning_rate": 0.00022058823529411765,
|
| 511 |
+
"loss": 0.9109,
|
| 512 |
+
"step": 5700
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 14.0,
|
| 516 |
+
"eval_loss": 0.9040451645851135,
|
| 517 |
+
"eval_runtime": 16.707,
|
| 518 |
+
"eval_samples_per_second": 478.841,
|
| 519 |
+
"eval_steps_per_second": 3.771,
|
| 520 |
+
"step": 5712
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 14.215686274509803,
|
| 524 |
+
"grad_norm": 14390.365234375,
|
| 525 |
+
"learning_rate": 0.00021568627450980395,
|
| 526 |
+
"loss": 0.8986,
|
| 527 |
+
"step": 5800
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 14.46078431372549,
|
| 531 |
+
"grad_norm": 12093.57421875,
|
| 532 |
+
"learning_rate": 0.00021078431372549021,
|
| 533 |
+
"loss": 0.9064,
|
| 534 |
+
"step": 5900
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 14.705882352941176,
|
| 538 |
+
"grad_norm": 9752.62890625,
|
| 539 |
+
"learning_rate": 0.00020588235294117645,
|
| 540 |
+
"loss": 0.9165,
|
| 541 |
+
"step": 6000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.950980392156863,
|
| 545 |
+
"grad_norm": 16142.40625,
|
| 546 |
+
"learning_rate": 0.00020098039215686275,
|
| 547 |
+
"loss": 0.9027,
|
| 548 |
+
"step": 6100
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 15.0,
|
| 552 |
+
"eval_loss": 0.9005336761474609,
|
| 553 |
+
"eval_runtime": 17.261,
|
| 554 |
+
"eval_samples_per_second": 463.471,
|
| 555 |
+
"eval_steps_per_second": 3.65,
|
| 556 |
+
"step": 6120
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 15.196078431372548,
|
| 560 |
+
"grad_norm": 14545.95703125,
|
| 561 |
+
"learning_rate": 0.000196078431372549,
|
| 562 |
+
"loss": 0.9078,
|
| 563 |
+
"step": 6200
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 15.441176470588236,
|
| 567 |
+
"grad_norm": 10779.703125,
|
| 568 |
+
"learning_rate": 0.00019117647058823528,
|
| 569 |
+
"loss": 0.8906,
|
| 570 |
+
"step": 6300
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 15.686274509803921,
|
| 574 |
+
"grad_norm": 11541.22265625,
|
| 575 |
+
"learning_rate": 0.00018627450980392157,
|
| 576 |
+
"loss": 0.8925,
|
| 577 |
+
"step": 6400
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 15.931372549019608,
|
| 581 |
+
"grad_norm": 11552.533203125,
|
| 582 |
+
"learning_rate": 0.00018137254901960784,
|
| 583 |
+
"loss": 0.8872,
|
| 584 |
+
"step": 6500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 16.0,
|
| 588 |
+
"eval_loss": 0.8934208154678345,
|
| 589 |
+
"eval_runtime": 16.6878,
|
| 590 |
+
"eval_samples_per_second": 479.393,
|
| 591 |
+
"eval_steps_per_second": 3.775,
|
| 592 |
+
"step": 6528
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 16.176470588235293,
|
| 596 |
+
"grad_norm": 11795.5029296875,
|
| 597 |
+
"learning_rate": 0.00017647058823529413,
|
| 598 |
+
"loss": 0.8854,
|
| 599 |
+
"step": 6600
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 16.42156862745098,
|
| 603 |
+
"grad_norm": 11742.291015625,
|
| 604 |
+
"learning_rate": 0.0001715686274509804,
|
| 605 |
+
"loss": 0.8853,
|
| 606 |
+
"step": 6700
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 16.666666666666668,
|
| 610 |
+
"grad_norm": 11037.279296875,
|
| 611 |
+
"learning_rate": 0.00016666666666666666,
|
| 612 |
+
"loss": 0.8805,
|
| 613 |
+
"step": 6800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 16.91176470588235,
|
| 617 |
+
"grad_norm": 13446.0478515625,
|
| 618 |
+
"learning_rate": 0.00016176470588235295,
|
| 619 |
+
"loss": 0.8754,
|
| 620 |
+
"step": 6900
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 17.0,
|
| 624 |
+
"eval_loss": 0.8874623775482178,
|
| 625 |
+
"eval_runtime": 16.6727,
|
| 626 |
+
"eval_samples_per_second": 479.827,
|
| 627 |
+
"eval_steps_per_second": 3.779,
|
| 628 |
+
"step": 6936
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 17.15686274509804,
|
| 632 |
+
"grad_norm": 13627.7470703125,
|
| 633 |
+
"learning_rate": 0.00015686274509803922,
|
| 634 |
+
"loss": 0.8801,
|
| 635 |
+
"step": 7000
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 17.401960784313726,
|
| 639 |
+
"grad_norm": 14940.4482421875,
|
| 640 |
+
"learning_rate": 0.00015196078431372549,
|
| 641 |
+
"loss": 0.8703,
|
| 642 |
+
"step": 7100
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 17.647058823529413,
|
| 646 |
+
"grad_norm": 17162.34765625,
|
| 647 |
+
"learning_rate": 0.00014705882352941178,
|
| 648 |
+
"loss": 0.8732,
|
| 649 |
+
"step": 7200
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 17.892156862745097,
|
| 653 |
+
"grad_norm": 13630.0888671875,
|
| 654 |
+
"learning_rate": 0.00014215686274509804,
|
| 655 |
+
"loss": 0.872,
|
| 656 |
+
"step": 7300
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 18.0,
|
| 660 |
+
"eval_loss": 0.8821218013763428,
|
| 661 |
+
"eval_runtime": 17.107,
|
| 662 |
+
"eval_samples_per_second": 467.645,
|
| 663 |
+
"eval_steps_per_second": 3.683,
|
| 664 |
+
"step": 7344
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 18.137254901960784,
|
| 668 |
+
"grad_norm": 12382.7177734375,
|
| 669 |
+
"learning_rate": 0.00013725490196078434,
|
| 670 |
+
"loss": 0.8609,
|
| 671 |
+
"step": 7400
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 18.38235294117647,
|
| 675 |
+
"grad_norm": 12689.1533203125,
|
| 676 |
+
"learning_rate": 0.0001323529411764706,
|
| 677 |
+
"loss": 0.8643,
|
| 678 |
+
"step": 7500
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 18.627450980392158,
|
| 682 |
+
"grad_norm": 13463.681640625,
|
| 683 |
+
"learning_rate": 0.00012745098039215687,
|
| 684 |
+
"loss": 0.8624,
|
| 685 |
+
"step": 7600
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 18.872549019607842,
|
| 689 |
+
"grad_norm": 11256.0576171875,
|
| 690 |
+
"learning_rate": 0.00012254901960784314,
|
| 691 |
+
"loss": 0.8648,
|
| 692 |
+
"step": 7700
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 19.0,
|
| 696 |
+
"eval_loss": 0.878471314907074,
|
| 697 |
+
"eval_runtime": 16.6776,
|
| 698 |
+
"eval_samples_per_second": 479.684,
|
| 699 |
+
"eval_steps_per_second": 3.778,
|
| 700 |
+
"step": 7752
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 19.11764705882353,
|
| 704 |
+
"grad_norm": 13279.96875,
|
| 705 |
+
"learning_rate": 0.00011764705882352942,
|
| 706 |
+
"loss": 0.8581,
|
| 707 |
+
"step": 7800
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 19.362745098039216,
|
| 711 |
+
"grad_norm": 12439.5205078125,
|
| 712 |
+
"learning_rate": 0.0001127450980392157,
|
| 713 |
+
"loss": 0.8624,
|
| 714 |
+
"step": 7900
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 19.607843137254903,
|
| 718 |
+
"grad_norm": 11728.5849609375,
|
| 719 |
+
"learning_rate": 0.00010784313725490197,
|
| 720 |
+
"loss": 0.8517,
|
| 721 |
+
"step": 8000
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 19.852941176470587,
|
| 725 |
+
"grad_norm": 11004.234375,
|
| 726 |
+
"learning_rate": 0.00010294117647058823,
|
| 727 |
+
"loss": 0.8476,
|
| 728 |
+
"step": 8100
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 20.0,
|
| 732 |
+
"eval_loss": 0.8746265769004822,
|
| 733 |
+
"eval_runtime": 16.7166,
|
| 734 |
+
"eval_samples_per_second": 478.566,
|
| 735 |
+
"eval_steps_per_second": 3.769,
|
| 736 |
+
"step": 8160
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 20.098039215686274,
|
| 740 |
+
"grad_norm": 12775.7421875,
|
| 741 |
+
"learning_rate": 9.80392156862745e-05,
|
| 742 |
+
"loss": 0.8636,
|
| 743 |
+
"step": 8200
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 20.34313725490196,
|
| 747 |
+
"grad_norm": 11157.79296875,
|
| 748 |
+
"learning_rate": 9.313725490196079e-05,
|
| 749 |
+
"loss": 0.8482,
|
| 750 |
+
"step": 8300
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 20.58823529411765,
|
| 754 |
+
"grad_norm": 12324.103515625,
|
| 755 |
+
"learning_rate": 8.823529411764706e-05,
|
| 756 |
+
"loss": 0.844,
|
| 757 |
+
"step": 8400
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 20.833333333333332,
|
| 761 |
+
"grad_norm": 12305.396484375,
|
| 762 |
+
"learning_rate": 8.333333333333333e-05,
|
| 763 |
+
"loss": 0.847,
|
| 764 |
+
"step": 8500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 21.0,
|
| 768 |
+
"eval_loss": 0.8715333342552185,
|
| 769 |
+
"eval_runtime": 16.7289,
|
| 770 |
+
"eval_samples_per_second": 478.213,
|
| 771 |
+
"eval_steps_per_second": 3.766,
|
| 772 |
+
"step": 8568
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 21.07843137254902,
|
| 776 |
+
"grad_norm": 14880.2978515625,
|
| 777 |
+
"learning_rate": 7.843137254901961e-05,
|
| 778 |
+
"loss": 0.8532,
|
| 779 |
+
"step": 8600
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 21.323529411764707,
|
| 783 |
+
"grad_norm": 11197.8076171875,
|
| 784 |
+
"learning_rate": 7.352941176470589e-05,
|
| 785 |
+
"loss": 0.8484,
|
| 786 |
+
"step": 8700
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 21.568627450980394,
|
| 790 |
+
"grad_norm": 10673.302734375,
|
| 791 |
+
"learning_rate": 6.862745098039217e-05,
|
| 792 |
+
"loss": 0.8395,
|
| 793 |
+
"step": 8800
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 21.813725490196077,
|
| 797 |
+
"grad_norm": 10516.341796875,
|
| 798 |
+
"learning_rate": 6.372549019607843e-05,
|
| 799 |
+
"loss": 0.8472,
|
| 800 |
+
"step": 8900
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 22.0,
|
| 804 |
+
"eval_loss": 0.8717733025550842,
|
| 805 |
+
"eval_runtime": 16.6953,
|
| 806 |
+
"eval_samples_per_second": 479.176,
|
| 807 |
+
"eval_steps_per_second": 3.774,
|
| 808 |
+
"step": 8976
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 22.058823529411764,
|
| 812 |
+
"grad_norm": 11651.111328125,
|
| 813 |
+
"learning_rate": 5.882352941176471e-05,
|
| 814 |
+
"loss": 0.8404,
|
| 815 |
+
"step": 9000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 22.30392156862745,
|
| 819 |
+
"grad_norm": 13144.53515625,
|
| 820 |
+
"learning_rate": 5.392156862745099e-05,
|
| 821 |
+
"loss": 0.8362,
|
| 822 |
+
"step": 9100
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 22.54901960784314,
|
| 826 |
+
"grad_norm": 11174.5419921875,
|
| 827 |
+
"learning_rate": 4.901960784313725e-05,
|
| 828 |
+
"loss": 0.8405,
|
| 829 |
+
"step": 9200
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 22.794117647058822,
|
| 833 |
+
"grad_norm": 10513.1904296875,
|
| 834 |
+
"learning_rate": 4.411764705882353e-05,
|
| 835 |
+
"loss": 0.8417,
|
| 836 |
+
"step": 9300
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 23.0,
|
| 840 |
+
"eval_loss": 0.8683518171310425,
|
| 841 |
+
"eval_runtime": 17.2108,
|
| 842 |
+
"eval_samples_per_second": 464.825,
|
| 843 |
+
"eval_steps_per_second": 3.66,
|
| 844 |
+
"step": 9384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 23.03921568627451,
|
| 848 |
+
"grad_norm": 13963.318359375,
|
| 849 |
+
"learning_rate": 3.9215686274509805e-05,
|
| 850 |
+
"loss": 0.8438,
|
| 851 |
+
"step": 9400
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 23.284313725490197,
|
| 855 |
+
"grad_norm": 15601.267578125,
|
| 856 |
+
"learning_rate": 3.4313725490196084e-05,
|
| 857 |
+
"loss": 0.8422,
|
| 858 |
+
"step": 9500
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 23.529411764705884,
|
| 862 |
+
"grad_norm": 12826.8515625,
|
| 863 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 864 |
+
"loss": 0.8304,
|
| 865 |
+
"step": 9600
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 23.774509803921568,
|
| 869 |
+
"grad_norm": 13277.5234375,
|
| 870 |
+
"learning_rate": 2.4509803921568626e-05,
|
| 871 |
+
"loss": 0.8357,
|
| 872 |
+
"step": 9700
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 24.0,
|
| 876 |
+
"eval_loss": 0.8688600659370422,
|
| 877 |
+
"eval_runtime": 16.7261,
|
| 878 |
+
"eval_samples_per_second": 478.293,
|
| 879 |
+
"eval_steps_per_second": 3.767,
|
| 880 |
+
"step": 9792
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 24.019607843137255,
|
| 884 |
+
"grad_norm": 11718.4677734375,
|
| 885 |
+
"learning_rate": 1.9607843137254903e-05,
|
| 886 |
+
"loss": 0.8361,
|
| 887 |
+
"step": 9800
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 24.264705882352942,
|
| 891 |
+
"grad_norm": 13389.1337890625,
|
| 892 |
+
"learning_rate": 1.4705882352941177e-05,
|
| 893 |
+
"loss": 0.8371,
|
| 894 |
+
"step": 9900
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 24.509803921568626,
|
| 898 |
+
"grad_norm": 10126.4189453125,
|
| 899 |
+
"learning_rate": 9.803921568627451e-06,
|
| 900 |
+
"loss": 0.8234,
|
| 901 |
+
"step": 10000
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 24.754901960784313,
|
| 905 |
+
"grad_norm": 15717.1884765625,
|
| 906 |
+
"learning_rate": 4.901960784313726e-06,
|
| 907 |
+
"loss": 0.8405,
|
| 908 |
+
"step": 10100
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 25.0,
|
| 912 |
+
"grad_norm": 27583.41015625,
|
| 913 |
+
"learning_rate": 0.0,
|
| 914 |
+
"loss": 0.8359,
|
| 915 |
+
"step": 10200
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 25.0,
|
| 919 |
+
"eval_loss": 0.8676381707191467,
|
| 920 |
+
"eval_runtime": 16.6913,
|
| 921 |
+
"eval_samples_per_second": 479.291,
|
| 922 |
+
"eval_steps_per_second": 3.774,
|
| 923 |
+
"step": 10200
|
| 924 |
+
}
|
| 925 |
+
],
|
| 926 |
+
"logging_steps": 100,
|
| 927 |
+
"max_steps": 10200,
|
| 928 |
+
"num_input_tokens_seen": 0,
|
| 929 |
+
"num_train_epochs": 25,
|
| 930 |
+
"save_steps": 500,
|
| 931 |
+
"stateful_callbacks": {
|
| 932 |
+
"EarlyStoppingCallback": {
|
| 933 |
+
"args": {
|
| 934 |
+
"early_stopping_patience": 2,
|
| 935 |
+
"early_stopping_threshold": 0.0
|
| 936 |
+
},
|
| 937 |
+
"attributes": {
|
| 938 |
+
"early_stopping_patience_counter": 0
|
| 939 |
+
}
|
| 940 |
+
},
|
| 941 |
+
"TrainerControl": {
|
| 942 |
+
"args": {
|
| 943 |
+
"should_epoch_stop": false,
|
| 944 |
+
"should_evaluate": false,
|
| 945 |
+
"should_log": false,
|
| 946 |
+
"should_save": true,
|
| 947 |
+
"should_training_stop": true
|
| 948 |
+
},
|
| 949 |
+
"attributes": {}
|
| 950 |
+
}
|
| 951 |
+
},
|
| 952 |
+
"total_flos": 4.40829394550784e+16,
|
| 953 |
+
"train_batch_size": 256,
|
| 954 |
+
"trial_name": null,
|
| 955 |
+
"trial_params": null
|
| 956 |
+
}
|
T5_ver3/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c6105412a6a505e5fdae2b12ff8d60ab2a484b094ffe9855d77224aff3e5360
|
| 3 |
+
size 5240
|
T5_ver3/vi_tokenizer_32128.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0fa3156da7d3526ca0cfe1f3c769ee3ca3c502737256934f00ed394a59389c85
|
| 3 |
+
size 748144
|