commit files to HF hub
Browse files- added_tokens.json +18 -0
- config.json +44 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +25 -0
- tokenizer.json +0 -0
- tokenizer_config.json +14 -0
- trainer_state.json +1036 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[ambiance-Negative]": 30526,
|
| 3 |
+
"[ambiance-Positive]": 30527,
|
| 4 |
+
"[ambiance-no majority]": 30529,
|
| 5 |
+
"[ambiance-unknown]": 30528,
|
| 6 |
+
"[food-Negative]": 30522,
|
| 7 |
+
"[food-Positive]": 30523,
|
| 8 |
+
"[food-no majority]": 30525,
|
| 9 |
+
"[food-unknown]": 30524,
|
| 10 |
+
"[noise-Negative]": 30530,
|
| 11 |
+
"[noise-Positive]": 30531,
|
| 12 |
+
"[noise-no majority]": 30533,
|
| 13 |
+
"[noise-unknown]": 30532,
|
| 14 |
+
"[service-Negative]": 30534,
|
| 15 |
+
"[service-Positive]": 30535,
|
| 16 |
+
"[service-no majority]": 30537,
|
| 17 |
+
"[service-unknown]": 30536
|
| 18 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "CEBaB/lstm.CEBaB.sa.5-class.exclusive.seed_42",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LSTMForFactualCounterfactualSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bidirectional": true,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"finetuning_task": "opentable",
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 300,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": 0,
|
| 16 |
+
"1": 1,
|
| 17 |
+
"2": 2,
|
| 18 |
+
"3": 3,
|
| 19 |
+
"4": 4
|
| 20 |
+
},
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"0": 0,
|
| 25 |
+
"1": 1,
|
| 26 |
+
"2": 2,
|
| 27 |
+
"3": 3,
|
| 28 |
+
"4": 4
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-12,
|
| 31 |
+
"max_position_embeddings": 512,
|
| 32 |
+
"model_type": "bert",
|
| 33 |
+
"num_attention_heads": 12,
|
| 34 |
+
"num_hidden_layers": 1,
|
| 35 |
+
"pad_token_id": 0,
|
| 36 |
+
"position_embedding_type": "absolute",
|
| 37 |
+
"problem_type": "single_label_classification",
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.21.1",
|
| 40 |
+
"type_vocab_size": 2,
|
| 41 |
+
"update_embeddings": false,
|
| 42 |
+
"use_cache": true,
|
| 43 |
+
"vocab_size": 30538
|
| 44 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80e8b1dbdd69c477b9e8a435debad8915a0bd56f5ac67626b3116f69c3bea383
|
| 3 |
+
size 85600319
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a1ccdf905ee3b6f2517bff7d55f38560f70065356fe0057c848f909729e0276
|
| 3 |
+
size 42799239
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62f55a84b23a7bee1dcd65623af2eb35c2f60acf55ab4688b927fed142207d78
|
| 3 |
+
size 14503
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5fb8a26c65dab7e59e0d28a5cebf95993ff70d3e93bb0996fd277196a51f038
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"[food-Negative]",
|
| 4 |
+
"[food-Positive]",
|
| 5 |
+
"[food-unknown]",
|
| 6 |
+
"[food-no majority]",
|
| 7 |
+
"[ambiance-Negative]",
|
| 8 |
+
"[ambiance-Positive]",
|
| 9 |
+
"[ambiance-unknown]",
|
| 10 |
+
"[ambiance-no majority]",
|
| 11 |
+
"[noise-Negative]",
|
| 12 |
+
"[noise-Positive]",
|
| 13 |
+
"[noise-unknown]",
|
| 14 |
+
"[noise-no majority]",
|
| 15 |
+
"[service-Negative]",
|
| 16 |
+
"[service-Positive]",
|
| 17 |
+
"[service-unknown]",
|
| 18 |
+
"[service-no majority]"
|
| 19 |
+
],
|
| 20 |
+
"cls_token": "[CLS]",
|
| 21 |
+
"mask_token": "[MASK]",
|
| 22 |
+
"pad_token": "[PAD]",
|
| 23 |
+
"sep_token": "[SEP]",
|
| 24 |
+
"unk_token": "[UNK]"
|
| 25 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"do_lower_case": true,
|
| 4 |
+
"mask_token": "[MASK]",
|
| 5 |
+
"model_max_length": 512,
|
| 6 |
+
"name_or_path": "bert-base-uncased",
|
| 7 |
+
"pad_token": "[PAD]",
|
| 8 |
+
"sep_token": "[SEP]",
|
| 9 |
+
"special_tokens_map_file": null,
|
| 10 |
+
"strip_accents": null,
|
| 11 |
+
"tokenize_chinese_chars": true,
|
| 12 |
+
"tokenizer_class": "BertTokenizer",
|
| 13 |
+
"unk_token": "[UNK]"
|
| 14 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1036 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5015928149223328,
|
| 3 |
+
"best_model_checkpoint": "model_output/e2e_opentable_5_way__approximate__0-shot__seed-42__lstm/checkpoint-1550",
|
| 4 |
+
"epoch": 2.6586620926243567,
|
| 5 |
+
"global_step": 1550,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 4.9783362218370885e-05,
|
| 13 |
+
"loss": 43.0786,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.07,
|
| 18 |
+
"learning_rate": 4.956672443674177e-05,
|
| 19 |
+
"loss": 38.7276,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"eval_distillation_accuracy_counterfactual": 0.413801949717804,
|
| 25 |
+
"eval_distillation_accuracy_factual": 0.7095946639302206,
|
| 26 |
+
"eval_distillation_f1_counterfactual": 0.3810686639321449,
|
| 27 |
+
"eval_distillation_f1_factual": 0.6957955686304648,
|
| 28 |
+
"eval_groundtruth_accuracy_counterfactual": 0.3301693175987686,
|
| 29 |
+
"eval_groundtruth_f1_counterfactual": 0.3009236688640019,
|
| 30 |
+
"eval_groundtruth_f1_factual": 0.4899999707848249,
|
| 31 |
+
"eval_icace_cosine": 0.7151280045509338,
|
| 32 |
+
"eval_icace_l2": 0.7787489891052246,
|
| 33 |
+
"eval_icace_normdiff": 0.5948067307472229,
|
| 34 |
+
"eval_loss": 20.85917091369629,
|
| 35 |
+
"eval_runtime": 4.2958,
|
| 36 |
+
"eval_samples_per_second": 907.39,
|
| 37 |
+
"eval_steps_per_second": 7.216,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1,
|
| 42 |
+
"learning_rate": 4.935008665511265e-05,
|
| 43 |
+
"loss": 35.9292,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 4.913344887348354e-05,
|
| 49 |
+
"loss": 32.7092,
|
| 50 |
+
"step": 80
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 4.891681109185442e-05,
|
| 55 |
+
"loss": 28.9099,
|
| 56 |
+
"step": 100
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.17,
|
| 60 |
+
"eval_distillation_accuracy_counterfactual": 0.44971780400205236,
|
| 61 |
+
"eval_distillation_accuracy_factual": 0.7447408927655208,
|
| 62 |
+
"eval_distillation_f1_counterfactual": 0.4113364356485835,
|
| 63 |
+
"eval_distillation_f1_factual": 0.7313179359518789,
|
| 64 |
+
"eval_groundtruth_accuracy_counterfactual": 0.3763468445356593,
|
| 65 |
+
"eval_groundtruth_f1_counterfactual": 0.340596840887459,
|
| 66 |
+
"eval_groundtruth_f1_factual": 0.5410055670992968,
|
| 67 |
+
"eval_icace_cosine": 0.5891422629356384,
|
| 68 |
+
"eval_icace_l2": 0.708361029624939,
|
| 69 |
+
"eval_icace_normdiff": 0.5622901916503906,
|
| 70 |
+
"eval_loss": 15.46045207977295,
|
| 71 |
+
"eval_runtime": 4.856,
|
| 72 |
+
"eval_samples_per_second": 802.712,
|
| 73 |
+
"eval_steps_per_second": 6.384,
|
| 74 |
+
"step": 100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 4.8700173310225307e-05,
|
| 79 |
+
"loss": 23.7929,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24,
|
| 84 |
+
"learning_rate": 4.848353552859619e-05,
|
| 85 |
+
"loss": 19.9476,
|
| 86 |
+
"step": 140
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26,
|
| 90 |
+
"eval_distillation_accuracy_counterfactual": 0.4807593637762955,
|
| 91 |
+
"eval_distillation_accuracy_factual": 0.6631605951770139,
|
| 92 |
+
"eval_distillation_f1_counterfactual": 0.4198039457002024,
|
| 93 |
+
"eval_distillation_f1_factual": 0.6325010855821678,
|
| 94 |
+
"eval_groundtruth_accuracy_counterfactual": 0.439712673165726,
|
| 95 |
+
"eval_groundtruth_f1_counterfactual": 0.38279540235888865,
|
| 96 |
+
"eval_groundtruth_f1_factual": 0.49479750987904253,
|
| 97 |
+
"eval_icace_cosine": 0.5389060378074646,
|
| 98 |
+
"eval_icace_l2": 0.6715835332870483,
|
| 99 |
+
"eval_icace_normdiff": 0.43847718834877014,
|
| 100 |
+
"eval_loss": 12.59195613861084,
|
| 101 |
+
"eval_runtime": 4.5282,
|
| 102 |
+
"eval_samples_per_second": 860.827,
|
| 103 |
+
"eval_steps_per_second": 6.846,
|
| 104 |
+
"step": 150
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.27,
|
| 108 |
+
"learning_rate": 4.826689774696707e-05,
|
| 109 |
+
"loss": 18.2834,
|
| 110 |
+
"step": 160
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.31,
|
| 114 |
+
"learning_rate": 4.8050259965337955e-05,
|
| 115 |
+
"loss": 16.7884,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.34,
|
| 120 |
+
"learning_rate": 4.7833622183708845e-05,
|
| 121 |
+
"loss": 16.1986,
|
| 122 |
+
"step": 200
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.34,
|
| 126 |
+
"eval_distillation_accuracy_counterfactual": 0.5100051308363264,
|
| 127 |
+
"eval_distillation_accuracy_factual": 0.7208825038481272,
|
| 128 |
+
"eval_distillation_f1_counterfactual": 0.44399257347283927,
|
| 129 |
+
"eval_distillation_f1_factual": 0.6985538094507235,
|
| 130 |
+
"eval_groundtruth_accuracy_counterfactual": 0.469728065674705,
|
| 131 |
+
"eval_groundtruth_f1_counterfactual": 0.4066620855800582,
|
| 132 |
+
"eval_groundtruth_f1_factual": 0.5453203881049435,
|
| 133 |
+
"eval_icace_cosine": 0.5247439742088318,
|
| 134 |
+
"eval_icace_l2": 0.6397933959960938,
|
| 135 |
+
"eval_icace_normdiff": 0.42328375577926636,
|
| 136 |
+
"eval_loss": 11.142946243286133,
|
| 137 |
+
"eval_runtime": 4.7099,
|
| 138 |
+
"eval_samples_per_second": 827.623,
|
| 139 |
+
"eval_steps_per_second": 6.582,
|
| 140 |
+
"step": 200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.38,
|
| 144 |
+
"learning_rate": 4.761698440207972e-05,
|
| 145 |
+
"loss": 16.3317,
|
| 146 |
+
"step": 220
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.41,
|
| 150 |
+
"learning_rate": 4.740034662045061e-05,
|
| 151 |
+
"loss": 15.6449,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.43,
|
| 156 |
+
"eval_distillation_accuracy_counterfactual": 0.5177013853258081,
|
| 157 |
+
"eval_distillation_accuracy_factual": 0.7614161108260646,
|
| 158 |
+
"eval_distillation_f1_counterfactual": 0.43564292507442814,
|
| 159 |
+
"eval_distillation_f1_factual": 0.7349550078954596,
|
| 160 |
+
"eval_groundtruth_accuracy_counterfactual": 0.48255515649050795,
|
| 161 |
+
"eval_groundtruth_f1_counterfactual": 0.40504065429982916,
|
| 162 |
+
"eval_groundtruth_f1_factual": 0.5535077645302229,
|
| 163 |
+
"eval_icace_cosine": 0.516343355178833,
|
| 164 |
+
"eval_icace_l2": 0.6303517818450928,
|
| 165 |
+
"eval_icace_normdiff": 0.4126652777194977,
|
| 166 |
+
"eval_loss": 10.60521411895752,
|
| 167 |
+
"eval_runtime": 4.617,
|
| 168 |
+
"eval_samples_per_second": 844.277,
|
| 169 |
+
"eval_steps_per_second": 6.714,
|
| 170 |
+
"step": 250
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45,
|
| 174 |
+
"learning_rate": 4.7183708838821494e-05,
|
| 175 |
+
"loss": 15.2981,
|
| 176 |
+
"step": 260
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.48,
|
| 180 |
+
"learning_rate": 4.6967071057192376e-05,
|
| 181 |
+
"loss": 15.1658,
|
| 182 |
+
"step": 280
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.51,
|
| 186 |
+
"learning_rate": 4.675043327556326e-05,
|
| 187 |
+
"loss": 15.0967,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.51,
|
| 192 |
+
"eval_distillation_accuracy_counterfactual": 0.5151359671626475,
|
| 193 |
+
"eval_distillation_accuracy_factual": 0.7678296562339662,
|
| 194 |
+
"eval_distillation_f1_counterfactual": 0.4566628678247276,
|
| 195 |
+
"eval_distillation_f1_factual": 0.7434229464297083,
|
| 196 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4856336582863007,
|
| 197 |
+
"eval_groundtruth_f1_counterfactual": 0.43108046151644946,
|
| 198 |
+
"eval_groundtruth_f1_factual": 0.5711126744289745,
|
| 199 |
+
"eval_icace_cosine": 0.5121855735778809,
|
| 200 |
+
"eval_icace_l2": 0.636752188205719,
|
| 201 |
+
"eval_icace_normdiff": 0.41414302587509155,
|
| 202 |
+
"eval_loss": 10.544304847717285,
|
| 203 |
+
"eval_runtime": 4.8047,
|
| 204 |
+
"eval_samples_per_second": 811.294,
|
| 205 |
+
"eval_steps_per_second": 6.452,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.55,
|
| 210 |
+
"learning_rate": 4.653379549393415e-05,
|
| 211 |
+
"loss": 14.9139,
|
| 212 |
+
"step": 320
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.58,
|
| 216 |
+
"learning_rate": 4.6317157712305025e-05,
|
| 217 |
+
"loss": 14.7833,
|
| 218 |
+
"step": 340
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.6,
|
| 222 |
+
"eval_distillation_accuracy_counterfactual": 0.525910723447922,
|
| 223 |
+
"eval_distillation_accuracy_factual": 0.7939969214982042,
|
| 224 |
+
"eval_distillation_f1_counterfactual": 0.4562979855469525,
|
| 225 |
+
"eval_distillation_f1_factual": 0.7748803987234086,
|
| 226 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5010261672652643,
|
| 227 |
+
"eval_groundtruth_f1_counterfactual": 0.43277911965474775,
|
| 228 |
+
"eval_groundtruth_f1_factual": 0.5673297128547324,
|
| 229 |
+
"eval_icace_cosine": 0.5058079957962036,
|
| 230 |
+
"eval_icace_l2": 0.6291938424110413,
|
| 231 |
+
"eval_icace_normdiff": 0.41149842739105225,
|
| 232 |
+
"eval_loss": 10.096619606018066,
|
| 233 |
+
"eval_runtime": 4.807,
|
| 234 |
+
"eval_samples_per_second": 810.901,
|
| 235 |
+
"eval_steps_per_second": 6.449,
|
| 236 |
+
"step": 350
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.62,
|
| 240 |
+
"learning_rate": 4.6100519930675915e-05,
|
| 241 |
+
"loss": 14.3751,
|
| 242 |
+
"step": 360
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.65,
|
| 246 |
+
"learning_rate": 4.58838821490468e-05,
|
| 247 |
+
"loss": 14.1618,
|
| 248 |
+
"step": 380
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.69,
|
| 252 |
+
"learning_rate": 4.566724436741768e-05,
|
| 253 |
+
"loss": 14.1163,
|
| 254 |
+
"step": 400
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.69,
|
| 258 |
+
"eval_distillation_accuracy_counterfactual": 0.5110312981015905,
|
| 259 |
+
"eval_distillation_accuracy_factual": 0.8142637249871729,
|
| 260 |
+
"eval_distillation_f1_counterfactual": 0.42866472439070946,
|
| 261 |
+
"eval_distillation_f1_factual": 0.7995337396770525,
|
| 262 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5002565418163161,
|
| 263 |
+
"eval_groundtruth_f1_counterfactual": 0.4196447531444125,
|
| 264 |
+
"eval_groundtruth_f1_factual": 0.570939084196488,
|
| 265 |
+
"eval_icace_cosine": 0.5057737827301025,
|
| 266 |
+
"eval_icace_l2": 0.6323451399803162,
|
| 267 |
+
"eval_icace_normdiff": 0.4111088812351227,
|
| 268 |
+
"eval_loss": 9.999621391296387,
|
| 269 |
+
"eval_runtime": 4.6656,
|
| 270 |
+
"eval_samples_per_second": 835.478,
|
| 271 |
+
"eval_steps_per_second": 6.644,
|
| 272 |
+
"step": 400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.72,
|
| 276 |
+
"learning_rate": 4.5450606585788563e-05,
|
| 277 |
+
"loss": 14.1471,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.75,
|
| 282 |
+
"learning_rate": 4.5233968804159446e-05,
|
| 283 |
+
"loss": 14.0263,
|
| 284 |
+
"step": 440
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.77,
|
| 288 |
+
"eval_distillation_accuracy_counterfactual": 0.5094920472036942,
|
| 289 |
+
"eval_distillation_accuracy_factual": 0.8224730631092868,
|
| 290 |
+
"eval_distillation_f1_counterfactual": 0.4472075955335491,
|
| 291 |
+
"eval_distillation_f1_factual": 0.8071684604194267,
|
| 292 |
+
"eval_groundtruth_accuracy_counterfactual": 0.49640841457157514,
|
| 293 |
+
"eval_groundtruth_f1_counterfactual": 0.43563795723268167,
|
| 294 |
+
"eval_groundtruth_f1_factual": 0.5741319757982798,
|
| 295 |
+
"eval_icace_cosine": 0.5053402185440063,
|
| 296 |
+
"eval_icace_l2": 0.6384401917457581,
|
| 297 |
+
"eval_icace_normdiff": 0.41610294580459595,
|
| 298 |
+
"eval_loss": 9.992120742797852,
|
| 299 |
+
"eval_runtime": 4.6252,
|
| 300 |
+
"eval_samples_per_second": 842.767,
|
| 301 |
+
"eval_steps_per_second": 6.702,
|
| 302 |
+
"step": 450
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.79,
|
| 306 |
+
"learning_rate": 4.501733102253033e-05,
|
| 307 |
+
"loss": 13.777,
|
| 308 |
+
"step": 460
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.82,
|
| 312 |
+
"learning_rate": 4.480069324090121e-05,
|
| 313 |
+
"loss": 13.7366,
|
| 314 |
+
"step": 480
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.86,
|
| 318 |
+
"learning_rate": 4.45840554592721e-05,
|
| 319 |
+
"loss": 13.645,
|
| 320 |
+
"step": 500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.86,
|
| 324 |
+
"eval_distillation_accuracy_counterfactual": 0.5187275525910724,
|
| 325 |
+
"eval_distillation_accuracy_factual": 0.8206772703950744,
|
| 326 |
+
"eval_distillation_f1_counterfactual": 0.44479298662465616,
|
| 327 |
+
"eval_distillation_f1_factual": 0.8084170023738808,
|
| 328 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5107747562852745,
|
| 329 |
+
"eval_groundtruth_f1_counterfactual": 0.43819547536098524,
|
| 330 |
+
"eval_groundtruth_f1_factual": 0.5804064040286399,
|
| 331 |
+
"eval_icace_cosine": 0.5102166533470154,
|
| 332 |
+
"eval_icace_l2": 0.6428956985473633,
|
| 333 |
+
"eval_icace_normdiff": 0.4216759502887726,
|
| 334 |
+
"eval_loss": 10.014728546142578,
|
| 335 |
+
"eval_runtime": 4.6422,
|
| 336 |
+
"eval_samples_per_second": 839.683,
|
| 337 |
+
"eval_steps_per_second": 6.678,
|
| 338 |
+
"step": 500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.89,
|
| 342 |
+
"learning_rate": 4.436741767764298e-05,
|
| 343 |
+
"loss": 13.6366,
|
| 344 |
+
"step": 520
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.93,
|
| 348 |
+
"learning_rate": 4.415077989601387e-05,
|
| 349 |
+
"loss": 13.5142,
|
| 350 |
+
"step": 540
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.94,
|
| 354 |
+
"eval_distillation_accuracy_counterfactual": 0.521292970754233,
|
| 355 |
+
"eval_distillation_accuracy_factual": 0.8222165212929707,
|
| 356 |
+
"eval_distillation_f1_counterfactual": 0.44615115849333264,
|
| 357 |
+
"eval_distillation_f1_factual": 0.8098912953108461,
|
| 358 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5138532580810672,
|
| 359 |
+
"eval_groundtruth_f1_counterfactual": 0.4412909861048945,
|
| 360 |
+
"eval_groundtruth_f1_factual": 0.5759446881541332,
|
| 361 |
+
"eval_icace_cosine": 0.5050559043884277,
|
| 362 |
+
"eval_icace_l2": 0.6343655586242676,
|
| 363 |
+
"eval_icace_normdiff": 0.4120153784751892,
|
| 364 |
+
"eval_loss": 9.914525032043457,
|
| 365 |
+
"eval_runtime": 4.5769,
|
| 366 |
+
"eval_samples_per_second": 851.659,
|
| 367 |
+
"eval_steps_per_second": 6.773,
|
| 368 |
+
"step": 550
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.96,
|
| 372 |
+
"learning_rate": 4.393414211438475e-05,
|
| 373 |
+
"loss": 13.516,
|
| 374 |
+
"step": 560
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.99,
|
| 378 |
+
"learning_rate": 4.371750433275563e-05,
|
| 379 |
+
"loss": 13.2886,
|
| 380 |
+
"step": 580
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.03,
|
| 384 |
+
"learning_rate": 4.3500866551126516e-05,
|
| 385 |
+
"loss": 13.0863,
|
| 386 |
+
"step": 600
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.03,
|
| 390 |
+
"eval_distillation_accuracy_counterfactual": 0.5225756798358132,
|
| 391 |
+
"eval_distillation_accuracy_factual": 0.8322216521292971,
|
| 392 |
+
"eval_distillation_f1_counterfactual": 0.44257027446874675,
|
| 393 |
+
"eval_distillation_f1_factual": 0.8222712281483833,
|
| 394 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5159055926115957,
|
| 395 |
+
"eval_groundtruth_f1_counterfactual": 0.43868456910889764,
|
| 396 |
+
"eval_groundtruth_f1_factual": 0.5752420543125346,
|
| 397 |
+
"eval_icace_cosine": 0.5034614205360413,
|
| 398 |
+
"eval_icace_l2": 0.6326570510864258,
|
| 399 |
+
"eval_icace_normdiff": 0.4119584262371063,
|
| 400 |
+
"eval_loss": 9.747785568237305,
|
| 401 |
+
"eval_runtime": 4.5826,
|
| 402 |
+
"eval_samples_per_second": 850.614,
|
| 403 |
+
"eval_steps_per_second": 6.765,
|
| 404 |
+
"step": 600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.06,
|
| 408 |
+
"learning_rate": 4.3284228769497406e-05,
|
| 409 |
+
"loss": 12.9339,
|
| 410 |
+
"step": 620
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.1,
|
| 414 |
+
"learning_rate": 4.306759098786828e-05,
|
| 415 |
+
"loss": 13.0796,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.11,
|
| 420 |
+
"eval_distillation_accuracy_counterfactual": 0.5110312981015905,
|
| 421 |
+
"eval_distillation_accuracy_factual": 0.8273473576192919,
|
| 422 |
+
"eval_distillation_f1_counterfactual": 0.43220510255327216,
|
| 423 |
+
"eval_distillation_f1_factual": 0.8195939186393728,
|
| 424 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5094920472036942,
|
| 425 |
+
"eval_groundtruth_f1_counterfactual": 0.43438648520235124,
|
| 426 |
+
"eval_groundtruth_f1_factual": 0.5765313785081062,
|
| 427 |
+
"eval_icace_cosine": 0.5092682242393494,
|
| 428 |
+
"eval_icace_l2": 0.6488710045814514,
|
| 429 |
+
"eval_icace_normdiff": 0.42473679780960083,
|
| 430 |
+
"eval_loss": 10.03084659576416,
|
| 431 |
+
"eval_runtime": 4.5757,
|
| 432 |
+
"eval_samples_per_second": 851.883,
|
| 433 |
+
"eval_steps_per_second": 6.775,
|
| 434 |
+
"step": 650
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.13,
|
| 438 |
+
"learning_rate": 4.285095320623917e-05,
|
| 439 |
+
"loss": 13.3328,
|
| 440 |
+
"step": 660
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.17,
|
| 444 |
+
"learning_rate": 4.2634315424610055e-05,
|
| 445 |
+
"loss": 13.2103,
|
| 446 |
+
"step": 680
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.2,
|
| 450 |
+
"learning_rate": 4.241767764298094e-05,
|
| 451 |
+
"loss": 12.8642,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.2,
|
| 456 |
+
"eval_distillation_accuracy_counterfactual": 0.5202668034889687,
|
| 457 |
+
"eval_distillation_accuracy_factual": 0.8227296049256029,
|
| 458 |
+
"eval_distillation_f1_counterfactual": 0.44726878993383945,
|
| 459 |
+
"eval_distillation_f1_factual": 0.8144807445603238,
|
| 460 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5146228835300154,
|
| 461 |
+
"eval_groundtruth_f1_counterfactual": 0.44463257796889355,
|
| 462 |
+
"eval_groundtruth_f1_factual": 0.5825450226028441,
|
| 463 |
+
"eval_icace_cosine": 0.5063404440879822,
|
| 464 |
+
"eval_icace_l2": 0.6389003396034241,
|
| 465 |
+
"eval_icace_normdiff": 0.41857677698135376,
|
| 466 |
+
"eval_loss": 9.888192176818848,
|
| 467 |
+
"eval_runtime": 4.6226,
|
| 468 |
+
"eval_samples_per_second": 843.25,
|
| 469 |
+
"eval_steps_per_second": 6.706,
|
| 470 |
+
"step": 700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.23,
|
| 474 |
+
"learning_rate": 4.220103986135182e-05,
|
| 475 |
+
"loss": 12.5476,
|
| 476 |
+
"step": 720
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.27,
|
| 480 |
+
"learning_rate": 4.198440207972271e-05,
|
| 481 |
+
"loss": 13.1858,
|
| 482 |
+
"step": 740
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.29,
|
| 486 |
+
"eval_distillation_accuracy_counterfactual": 0.5243714725500257,
|
| 487 |
+
"eval_distillation_accuracy_factual": 0.8327347357619292,
|
| 488 |
+
"eval_distillation_f1_counterfactual": 0.4460710855321577,
|
| 489 |
+
"eval_distillation_f1_factual": 0.8222063551634736,
|
| 490 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5182144689584403,
|
| 491 |
+
"eval_groundtruth_f1_counterfactual": 0.4434489043121113,
|
| 492 |
+
"eval_groundtruth_f1_factual": 0.5879251850701077,
|
| 493 |
+
"eval_icace_cosine": 0.5027849674224854,
|
| 494 |
+
"eval_icace_l2": 0.6359062790870667,
|
| 495 |
+
"eval_icace_normdiff": 0.41075506806373596,
|
| 496 |
+
"eval_loss": 9.917060852050781,
|
| 497 |
+
"eval_runtime": 4.7653,
|
| 498 |
+
"eval_samples_per_second": 817.997,
|
| 499 |
+
"eval_steps_per_second": 6.505,
|
| 500 |
+
"step": 750
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3,
|
| 504 |
+
"learning_rate": 4.1767764298093586e-05,
|
| 505 |
+
"loss": 12.7591,
|
| 506 |
+
"step": 760
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.34,
|
| 510 |
+
"learning_rate": 4.1551126516464476e-05,
|
| 511 |
+
"loss": 12.8519,
|
| 512 |
+
"step": 780
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.37,
|
| 516 |
+
"learning_rate": 4.133448873483536e-05,
|
| 517 |
+
"loss": 12.7851,
|
| 518 |
+
"step": 800
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.37,
|
| 522 |
+
"eval_distillation_accuracy_counterfactual": 0.5118009235505387,
|
| 523 |
+
"eval_distillation_accuracy_factual": 0.8137506413545408,
|
| 524 |
+
"eval_distillation_f1_counterfactual": 0.43914802457549407,
|
| 525 |
+
"eval_distillation_f1_factual": 0.8049341794937319,
|
| 526 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5079527963057978,
|
| 527 |
+
"eval_groundtruth_f1_counterfactual": 0.4407143465132375,
|
| 528 |
+
"eval_groundtruth_f1_factual": 0.5835148341351625,
|
| 529 |
+
"eval_icace_cosine": 0.5061963796615601,
|
| 530 |
+
"eval_icace_l2": 0.6484118103981018,
|
| 531 |
+
"eval_icace_normdiff": 0.42115285992622375,
|
| 532 |
+
"eval_loss": 10.143532752990723,
|
| 533 |
+
"eval_runtime": 4.6865,
|
| 534 |
+
"eval_samples_per_second": 831.75,
|
| 535 |
+
"eval_steps_per_second": 6.615,
|
| 536 |
+
"step": 800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.41,
|
| 540 |
+
"learning_rate": 4.111785095320624e-05,
|
| 541 |
+
"loss": 12.8646,
|
| 542 |
+
"step": 820
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.44,
|
| 546 |
+
"learning_rate": 4.0901213171577124e-05,
|
| 547 |
+
"loss": 12.558,
|
| 548 |
+
"step": 840
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.46,
|
| 552 |
+
"eval_distillation_accuracy_counterfactual": 0.5307850179579271,
|
| 553 |
+
"eval_distillation_accuracy_factual": 0.831965110312981,
|
| 554 |
+
"eval_distillation_f1_counterfactual": 0.4484519611985586,
|
| 555 |
+
"eval_distillation_f1_factual": 0.822168247394948,
|
| 556 |
+
"eval_groundtruth_accuracy_counterfactual": 0.517444843509492,
|
| 557 |
+
"eval_groundtruth_f1_counterfactual": 0.43872302047817147,
|
| 558 |
+
"eval_groundtruth_f1_factual": 0.5800644441761371,
|
| 559 |
+
"eval_icace_cosine": 0.5030389428138733,
|
| 560 |
+
"eval_icace_l2": 0.632723867893219,
|
| 561 |
+
"eval_icace_normdiff": 0.4054679870605469,
|
| 562 |
+
"eval_loss": 9.84372615814209,
|
| 563 |
+
"eval_runtime": 5.0033,
|
| 564 |
+
"eval_samples_per_second": 779.083,
|
| 565 |
+
"eval_steps_per_second": 6.196,
|
| 566 |
+
"step": 850
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.48,
|
| 570 |
+
"learning_rate": 4.068457538994801e-05,
|
| 571 |
+
"loss": 12.828,
|
| 572 |
+
"step": 860
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.51,
|
| 576 |
+
"learning_rate": 4.04679376083189e-05,
|
| 577 |
+
"loss": 12.3635,
|
| 578 |
+
"step": 880
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.54,
|
| 582 |
+
"learning_rate": 4.025129982668977e-05,
|
| 583 |
+
"loss": 12.6661,
|
| 584 |
+
"step": 900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.54,
|
| 588 |
+
"eval_distillation_accuracy_counterfactual": 0.5253976398152899,
|
| 589 |
+
"eval_distillation_accuracy_factual": 0.8365828630066701,
|
| 590 |
+
"eval_distillation_f1_counterfactual": 0.450312280094982,
|
| 591 |
+
"eval_distillation_f1_factual": 0.8262625380550788,
|
| 592 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5189840944073884,
|
| 593 |
+
"eval_groundtruth_f1_counterfactual": 0.44867654499260745,
|
| 594 |
+
"eval_groundtruth_f1_factual": 0.5750955137521894,
|
| 595 |
+
"eval_icace_cosine": 0.5052414536476135,
|
| 596 |
+
"eval_icace_l2": 0.6440022587776184,
|
| 597 |
+
"eval_icace_normdiff": 0.41912850737571716,
|
| 598 |
+
"eval_loss": 9.941903114318848,
|
| 599 |
+
"eval_runtime": 4.5789,
|
| 600 |
+
"eval_samples_per_second": 851.288,
|
| 601 |
+
"eval_steps_per_second": 6.77,
|
| 602 |
+
"step": 900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.58,
|
| 606 |
+
"learning_rate": 4.003466204506066e-05,
|
| 607 |
+
"loss": 12.9309,
|
| 608 |
+
"step": 920
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.61,
|
| 612 |
+
"learning_rate": 3.9818024263431546e-05,
|
| 613 |
+
"loss": 12.463,
|
| 614 |
+
"step": 940
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.63,
|
| 618 |
+
"eval_distillation_accuracy_counterfactual": 0.513340174448435,
|
| 619 |
+
"eval_distillation_accuracy_factual": 0.8099025141097999,
|
| 620 |
+
"eval_distillation_f1_counterfactual": 0.4451801537186667,
|
| 621 |
+
"eval_distillation_f1_factual": 0.8018083383301378,
|
| 622 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5135967162647511,
|
| 623 |
+
"eval_groundtruth_f1_counterfactual": 0.45122773824298,
|
| 624 |
+
"eval_groundtruth_f1_factual": 0.582459738995094,
|
| 625 |
+
"eval_icace_cosine": 0.5057380199432373,
|
| 626 |
+
"eval_icace_l2": 0.6424580216407776,
|
| 627 |
+
"eval_icace_normdiff": 0.4136655926704407,
|
| 628 |
+
"eval_loss": 10.070512771606445,
|
| 629 |
+
"eval_runtime": 4.5868,
|
| 630 |
+
"eval_samples_per_second": 849.823,
|
| 631 |
+
"eval_steps_per_second": 6.758,
|
| 632 |
+
"step": 950
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.65,
|
| 636 |
+
"learning_rate": 3.960138648180243e-05,
|
| 637 |
+
"loss": 12.6212,
|
| 638 |
+
"step": 960
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.68,
|
| 642 |
+
"learning_rate": 3.938474870017331e-05,
|
| 643 |
+
"loss": 12.6221,
|
| 644 |
+
"step": 980
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.72,
|
| 648 |
+
"learning_rate": 3.91681109185442e-05,
|
| 649 |
+
"loss": 12.5131,
|
| 650 |
+
"step": 1000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.72,
|
| 654 |
+
"eval_distillation_accuracy_counterfactual": 0.5192406362237044,
|
| 655 |
+
"eval_distillation_accuracy_factual": 0.8127244740892765,
|
| 656 |
+
"eval_distillation_f1_counterfactual": 0.4342093180083893,
|
| 657 |
+
"eval_distillation_f1_factual": 0.8020837833036161,
|
| 658 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5156490507952797,
|
| 659 |
+
"eval_groundtruth_f1_counterfactual": 0.4371545167935388,
|
| 660 |
+
"eval_groundtruth_f1_factual": 0.5865676169660375,
|
| 661 |
+
"eval_icace_cosine": 0.5021520256996155,
|
| 662 |
+
"eval_icace_l2": 0.6322320699691772,
|
| 663 |
+
"eval_icace_normdiff": 0.40523087978363037,
|
| 664 |
+
"eval_loss": 9.895723342895508,
|
| 665 |
+
"eval_runtime": 4.6904,
|
| 666 |
+
"eval_samples_per_second": 831.068,
|
| 667 |
+
"eval_steps_per_second": 6.609,
|
| 668 |
+
"step": 1000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.75,
|
| 672 |
+
"learning_rate": 3.895147313691508e-05,
|
| 673 |
+
"loss": 12.5027,
|
| 674 |
+
"step": 1020
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.78,
|
| 678 |
+
"learning_rate": 3.873483535528597e-05,
|
| 679 |
+
"loss": 12.7083,
|
| 680 |
+
"step": 1040
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.8,
|
| 684 |
+
"eval_distillation_accuracy_counterfactual": 0.5192406362237044,
|
| 685 |
+
"eval_distillation_accuracy_factual": 0.8114417650076963,
|
| 686 |
+
"eval_distillation_f1_counterfactual": 0.4263386972710933,
|
| 687 |
+
"eval_distillation_f1_factual": 0.7994083878081278,
|
| 688 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5192406362237044,
|
| 689 |
+
"eval_groundtruth_f1_counterfactual": 0.4329553891077026,
|
| 690 |
+
"eval_groundtruth_f1_factual": 0.5912280932648739,
|
| 691 |
+
"eval_icace_cosine": 0.5038776397705078,
|
| 692 |
+
"eval_icace_l2": 0.6317997574806213,
|
| 693 |
+
"eval_icace_normdiff": 0.39978867769241333,
|
| 694 |
+
"eval_loss": 10.077560424804688,
|
| 695 |
+
"eval_runtime": 5.0771,
|
| 696 |
+
"eval_samples_per_second": 767.767,
|
| 697 |
+
"eval_steps_per_second": 6.106,
|
| 698 |
+
"step": 1050
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.82,
|
| 702 |
+
"learning_rate": 3.851819757365685e-05,
|
| 703 |
+
"loss": 12.3779,
|
| 704 |
+
"step": 1060
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.85,
|
| 708 |
+
"learning_rate": 3.830155979202773e-05,
|
| 709 |
+
"loss": 12.1297,
|
| 710 |
+
"step": 1080
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.89,
|
| 714 |
+
"learning_rate": 3.8084922010398616e-05,
|
| 715 |
+
"loss": 12.2778,
|
| 716 |
+
"step": 1100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.89,
|
| 720 |
+
"eval_distillation_accuracy_counterfactual": 0.5076962544894817,
|
| 721 |
+
"eval_distillation_accuracy_factual": 0.8093894304771678,
|
| 722 |
+
"eval_distillation_f1_counterfactual": 0.42915135801638715,
|
| 723 |
+
"eval_distillation_f1_factual": 0.7988387328754103,
|
| 724 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5074397126731657,
|
| 725 |
+
"eval_groundtruth_f1_counterfactual": 0.43467229220494497,
|
| 726 |
+
"eval_groundtruth_f1_factual": 0.5863614417021632,
|
| 727 |
+
"eval_icace_cosine": 0.5077196955680847,
|
| 728 |
+
"eval_icace_l2": 0.6488126516342163,
|
| 729 |
+
"eval_icace_normdiff": 0.4152773916721344,
|
| 730 |
+
"eval_loss": 10.29930591583252,
|
| 731 |
+
"eval_runtime": 4.7917,
|
| 732 |
+
"eval_samples_per_second": 813.496,
|
| 733 |
+
"eval_steps_per_second": 6.47,
|
| 734 |
+
"step": 1100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.92,
|
| 738 |
+
"learning_rate": 3.78682842287695e-05,
|
| 739 |
+
"loss": 12.391,
|
| 740 |
+
"step": 1120
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.96,
|
| 744 |
+
"learning_rate": 3.765164644714038e-05,
|
| 745 |
+
"loss": 12.1049,
|
| 746 |
+
"step": 1140
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.97,
|
| 750 |
+
"eval_distillation_accuracy_counterfactual": 0.5277065161621344,
|
| 751 |
+
"eval_distillation_accuracy_factual": 0.8114417650076963,
|
| 752 |
+
"eval_distillation_f1_counterfactual": 0.4379402413535344,
|
| 753 |
+
"eval_distillation_f1_factual": 0.8009596923401426,
|
| 754 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5184710107747563,
|
| 755 |
+
"eval_groundtruth_f1_counterfactual": 0.4351158243705179,
|
| 756 |
+
"eval_groundtruth_f1_factual": 0.5809362022910343,
|
| 757 |
+
"eval_icace_cosine": 0.5041907429695129,
|
| 758 |
+
"eval_icace_l2": 0.6319468021392822,
|
| 759 |
+
"eval_icace_normdiff": 0.4021805226802826,
|
| 760 |
+
"eval_loss": 10.031487464904785,
|
| 761 |
+
"eval_runtime": 4.6011,
|
| 762 |
+
"eval_samples_per_second": 847.186,
|
| 763 |
+
"eval_steps_per_second": 6.737,
|
| 764 |
+
"step": 1150
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.99,
|
| 768 |
+
"learning_rate": 3.7435008665511264e-05,
|
| 769 |
+
"loss": 12.3362,
|
| 770 |
+
"step": 1160
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.02,
|
| 774 |
+
"learning_rate": 3.7218370883882154e-05,
|
| 775 |
+
"loss": 12.3522,
|
| 776 |
+
"step": 1180
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.06,
|
| 780 |
+
"learning_rate": 3.700173310225303e-05,
|
| 781 |
+
"loss": 12.2852,
|
| 782 |
+
"step": 1200
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.06,
|
| 786 |
+
"eval_distillation_accuracy_counterfactual": 0.5089789635710621,
|
| 787 |
+
"eval_distillation_accuracy_factual": 0.8093894304771678,
|
| 788 |
+
"eval_distillation_f1_counterfactual": 0.4327873004366438,
|
| 789 |
+
"eval_distillation_f1_factual": 0.7978534612810504,
|
| 790 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5094920472036942,
|
| 791 |
+
"eval_groundtruth_f1_counterfactual": 0.44147489438032766,
|
| 792 |
+
"eval_groundtruth_f1_factual": 0.5875097166633956,
|
| 793 |
+
"eval_icace_cosine": 0.5080219507217407,
|
| 794 |
+
"eval_icace_l2": 0.6420192122459412,
|
| 795 |
+
"eval_icace_normdiff": 0.40897807478904724,
|
| 796 |
+
"eval_loss": 10.132452011108398,
|
| 797 |
+
"eval_runtime": 4.7301,
|
| 798 |
+
"eval_samples_per_second": 824.088,
|
| 799 |
+
"eval_steps_per_second": 6.554,
|
| 800 |
+
"step": 1200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.09,
|
| 804 |
+
"learning_rate": 3.678509532062392e-05,
|
| 805 |
+
"loss": 12.2919,
|
| 806 |
+
"step": 1220
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.13,
|
| 810 |
+
"learning_rate": 3.65684575389948e-05,
|
| 811 |
+
"loss": 12.2076,
|
| 812 |
+
"step": 1240
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.14,
|
| 816 |
+
"eval_distillation_accuracy_counterfactual": 0.5082093381221139,
|
| 817 |
+
"eval_distillation_accuracy_factual": 0.8052847614161108,
|
| 818 |
+
"eval_distillation_f1_counterfactual": 0.421327115393484,
|
| 819 |
+
"eval_distillation_f1_factual": 0.7938972956790581,
|
| 820 |
+
"eval_groundtruth_accuracy_counterfactual": 0.508722421754746,
|
| 821 |
+
"eval_groundtruth_f1_counterfactual": 0.4317767637232814,
|
| 822 |
+
"eval_groundtruth_f1_factual": 0.5884182312452589,
|
| 823 |
+
"eval_icace_cosine": 0.5102863907814026,
|
| 824 |
+
"eval_icace_l2": 0.6493218541145325,
|
| 825 |
+
"eval_icace_normdiff": 0.41481640934944153,
|
| 826 |
+
"eval_loss": 10.397825241088867,
|
| 827 |
+
"eval_runtime": 4.5642,
|
| 828 |
+
"eval_samples_per_second": 854.035,
|
| 829 |
+
"eval_steps_per_second": 6.792,
|
| 830 |
+
"step": 1250
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.16,
|
| 834 |
+
"learning_rate": 3.6351819757365686e-05,
|
| 835 |
+
"loss": 12.1793,
|
| 836 |
+
"step": 1260
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2,
|
| 840 |
+
"learning_rate": 3.613518197573657e-05,
|
| 841 |
+
"loss": 11.9955,
|
| 842 |
+
"step": 1280
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.23,
|
| 846 |
+
"learning_rate": 3.591854419410746e-05,
|
| 847 |
+
"loss": 12.0414,
|
| 848 |
+
"step": 1300
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.23,
|
| 852 |
+
"eval_distillation_accuracy_counterfactual": 0.5159055926115957,
|
| 853 |
+
"eval_distillation_accuracy_factual": 0.8088763468445357,
|
| 854 |
+
"eval_distillation_f1_counterfactual": 0.43928335663414925,
|
| 855 |
+
"eval_distillation_f1_factual": 0.798007092912107,
|
| 856 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5192406362237044,
|
| 857 |
+
"eval_groundtruth_f1_counterfactual": 0.45071084200597555,
|
| 858 |
+
"eval_groundtruth_f1_factual": 0.5881302910281572,
|
| 859 |
+
"eval_icace_cosine": 0.5018934607505798,
|
| 860 |
+
"eval_icace_l2": 0.6314676403999329,
|
| 861 |
+
"eval_icace_normdiff": 0.4030630588531494,
|
| 862 |
+
"eval_loss": 10.04751968383789,
|
| 863 |
+
"eval_runtime": 4.5717,
|
| 864 |
+
"eval_samples_per_second": 852.63,
|
| 865 |
+
"eval_steps_per_second": 6.781,
|
| 866 |
+
"step": 1300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.26,
|
| 870 |
+
"learning_rate": 3.5701906412478334e-05,
|
| 871 |
+
"loss": 12.0019,
|
| 872 |
+
"step": 1320
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.3,
|
| 876 |
+
"learning_rate": 3.5485268630849224e-05,
|
| 877 |
+
"loss": 11.8994,
|
| 878 |
+
"step": 1340
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.32,
|
| 882 |
+
"eval_distillation_accuracy_counterfactual": 0.5120574653668548,
|
| 883 |
+
"eval_distillation_accuracy_factual": 0.8001539250897897,
|
| 884 |
+
"eval_distillation_f1_counterfactual": 0.43648655334155206,
|
| 885 |
+
"eval_distillation_f1_factual": 0.791237363619484,
|
| 886 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5035915854284249,
|
| 887 |
+
"eval_groundtruth_f1_counterfactual": 0.4365541899982916,
|
| 888 |
+
"eval_groundtruth_f1_factual": 0.5892108143561166,
|
| 889 |
+
"eval_icace_cosine": 0.5103155970573425,
|
| 890 |
+
"eval_icace_l2": 0.648665726184845,
|
| 891 |
+
"eval_icace_normdiff": 0.4152573347091675,
|
| 892 |
+
"eval_loss": 10.379837989807129,
|
| 893 |
+
"eval_runtime": 4.5419,
|
| 894 |
+
"eval_samples_per_second": 858.233,
|
| 895 |
+
"eval_steps_per_second": 6.825,
|
| 896 |
+
"step": 1350
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.33,
|
| 900 |
+
"learning_rate": 3.526863084922011e-05,
|
| 901 |
+
"loss": 12.1835,
|
| 902 |
+
"step": 1360
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.37,
|
| 906 |
+
"learning_rate": 3.505199306759099e-05,
|
| 907 |
+
"loss": 12.1258,
|
| 908 |
+
"step": 1380
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.4,
|
| 912 |
+
"learning_rate": 3.483535528596187e-05,
|
| 913 |
+
"loss": 11.8758,
|
| 914 |
+
"step": 1400
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.4,
|
| 918 |
+
"eval_distillation_accuracy_counterfactual": 0.5092355053873782,
|
| 919 |
+
"eval_distillation_accuracy_factual": 0.8055413032324269,
|
| 920 |
+
"eval_distillation_f1_counterfactual": 0.43572512517349055,
|
| 921 |
+
"eval_distillation_f1_factual": 0.7959903986995227,
|
| 922 |
+
"eval_groundtruth_accuracy_counterfactual": 0.512827090815803,
|
| 923 |
+
"eval_groundtruth_f1_counterfactual": 0.44567177435185246,
|
| 924 |
+
"eval_groundtruth_f1_factual": 0.5944081922689601,
|
| 925 |
+
"eval_icace_cosine": 0.504929780960083,
|
| 926 |
+
"eval_icace_l2": 0.6384508013725281,
|
| 927 |
+
"eval_icace_normdiff": 0.404530793428421,
|
| 928 |
+
"eval_loss": 10.173859596252441,
|
| 929 |
+
"eval_runtime": 4.614,
|
| 930 |
+
"eval_samples_per_second": 844.817,
|
| 931 |
+
"eval_steps_per_second": 6.719,
|
| 932 |
+
"step": 1400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.44,
|
| 936 |
+
"learning_rate": 3.461871750433276e-05,
|
| 937 |
+
"loss": 12.0309,
|
| 938 |
+
"step": 1420
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.47,
|
| 942 |
+
"learning_rate": 3.440207972270364e-05,
|
| 943 |
+
"loss": 11.738,
|
| 944 |
+
"step": 1440
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.49,
|
| 948 |
+
"eval_distillation_accuracy_counterfactual": 0.5151359671626475,
|
| 949 |
+
"eval_distillation_accuracy_factual": 0.8065674704976912,
|
| 950 |
+
"eval_distillation_f1_counterfactual": 0.42557736492521486,
|
| 951 |
+
"eval_distillation_f1_factual": 0.7962593487138783,
|
| 952 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5207798871216008,
|
| 953 |
+
"eval_groundtruth_f1_counterfactual": 0.4400972677897436,
|
| 954 |
+
"eval_groundtruth_f1_factual": 0.5898699262629978,
|
| 955 |
+
"eval_icace_cosine": 0.504587709903717,
|
| 956 |
+
"eval_icace_l2": 0.6404756903648376,
|
| 957 |
+
"eval_icace_normdiff": 0.4062090516090393,
|
| 958 |
+
"eval_loss": 10.268754005432129,
|
| 959 |
+
"eval_runtime": 4.5043,
|
| 960 |
+
"eval_samples_per_second": 865.395,
|
| 961 |
+
"eval_steps_per_second": 6.882,
|
| 962 |
+
"step": 1450
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.5,
|
| 966 |
+
"learning_rate": 3.418544194107453e-05,
|
| 967 |
+
"loss": 11.847,
|
| 968 |
+
"step": 1460
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.54,
|
| 972 |
+
"learning_rate": 3.396880415944541e-05,
|
| 973 |
+
"loss": 11.7855,
|
| 974 |
+
"step": 1480
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.57,
|
| 978 |
+
"learning_rate": 3.3752166377816294e-05,
|
| 979 |
+
"loss": 12.2395,
|
| 980 |
+
"step": 1500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.57,
|
| 984 |
+
"eval_distillation_accuracy_counterfactual": 0.5076962544894817,
|
| 985 |
+
"eval_distillation_accuracy_factual": 0.793227296049256,
|
| 986 |
+
"eval_distillation_f1_counterfactual": 0.42559353759463275,
|
| 987 |
+
"eval_distillation_f1_factual": 0.7832657828168742,
|
| 988 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5105182144689584,
|
| 989 |
+
"eval_groundtruth_f1_counterfactual": 0.4380208778877491,
|
| 990 |
+
"eval_groundtruth_f1_factual": 0.5829268254157943,
|
| 991 |
+
"eval_icace_cosine": 0.508986234664917,
|
| 992 |
+
"eval_icace_l2": 0.6547452211380005,
|
| 993 |
+
"eval_icace_normdiff": 0.4193916618824005,
|
| 994 |
+
"eval_loss": 10.556840896606445,
|
| 995 |
+
"eval_runtime": 4.5426,
|
| 996 |
+
"eval_samples_per_second": 858.107,
|
| 997 |
+
"eval_steps_per_second": 6.824,
|
| 998 |
+
"step": 1500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.61,
|
| 1002 |
+
"learning_rate": 3.353552859618718e-05,
|
| 1003 |
+
"loss": 12.2207,
|
| 1004 |
+
"step": 1520
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.64,
|
| 1008 |
+
"learning_rate": 3.331889081455806e-05,
|
| 1009 |
+
"loss": 12.2085,
|
| 1010 |
+
"step": 1540
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.66,
|
| 1014 |
+
"eval_distillation_accuracy_counterfactual": 0.5177013853258081,
|
| 1015 |
+
"eval_distillation_accuracy_factual": 0.7978450487429452,
|
| 1016 |
+
"eval_distillation_f1_counterfactual": 0.4343412617259876,
|
| 1017 |
+
"eval_distillation_f1_factual": 0.7885736470487629,
|
| 1018 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5241149307337096,
|
| 1019 |
+
"eval_groundtruth_f1_counterfactual": 0.44795498771470293,
|
| 1020 |
+
"eval_groundtruth_f1_factual": 0.5848176602353289,
|
| 1021 |
+
"eval_icace_cosine": 0.5015928149223328,
|
| 1022 |
+
"eval_icace_l2": 0.6313996911048889,
|
| 1023 |
+
"eval_icace_normdiff": 0.400736927986145,
|
| 1024 |
+
"eval_loss": 10.176055908203125,
|
| 1025 |
+
"eval_runtime": 6.0094,
|
| 1026 |
+
"eval_samples_per_second": 648.649,
|
| 1027 |
+
"eval_steps_per_second": 5.159,
|
| 1028 |
+
"step": 1550
|
| 1029 |
+
}
|
| 1030 |
+
],
|
| 1031 |
+
"max_steps": 4616,
|
| 1032 |
+
"num_train_epochs": 8,
|
| 1033 |
+
"total_flos": 234153553766400.0,
|
| 1034 |
+
"trial_name": null,
|
| 1035 |
+
"trial_params": null
|
| 1036 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc4670877b5639566563026546ce1238d69367b163c4b03acbd6667c81a53688
|
| 3 |
+
size 3439
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|