commit files to HF hub
Browse files- added_tokens.json +1 -0
- config.json +43 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +1600 -0
- training_args.bin +3 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"[noise-unknown]": 50275, "[ambiance-Negative]": 50269, "[noise-no majority]": 50276, "[food-Negative]": 50265, "[food-Positive]": 50266, "[noise-Negative]": 50273, "[service-no majority]": 50280, "[ambiance-no majority]": 50272, "[service-Positive]": 50278, "[service-unknown]": 50279, "[ambiance-unknown]": 50271, "[service-Negative]": 50277, "[food-no majority]": 50268, "[noise-Positive]": 50274, "[ambiance-Positive]": 50270, "[food-unknown]": 50267}
|
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "CEBaB/roberta-base.CEBaB.sa.5-class.exclusive.seed_66",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForFactualCounterfactualSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"finetuning_task": "opentable",
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 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-05,
|
| 31 |
+
"max_position_embeddings": 514,
|
| 32 |
+
"model_type": "roberta",
|
| 33 |
+
"num_attention_heads": 12,
|
| 34 |
+
"num_hidden_layers": 12,
|
| 35 |
+
"pad_token_id": 1,
|
| 36 |
+
"position_embedding_type": "absolute",
|
| 37 |
+
"problem_type": "single_label_classification",
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.18.0",
|
| 40 |
+
"type_vocab_size": 1,
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 50281
|
| 43 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd29d2ce58aaac53d5e923008fd2b32ec41386df02a2e48cb4961dd75c8bf14a
|
| 3 |
+
size 997410589
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5daf33eaa573e0ba429f59d8341e658a10504db5428dae731a34a765e7a07256
|
| 3 |
+
size 498718701
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfed883d023e3f29d088a059f221c962a9819e146fa404132d1486e925da7de2
|
| 3 |
+
size 15523
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f3e71bbe8f880a1567f9f5e8a3bbbb746f675a180bf9b0fb1ff897735641a7f
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}, "additional_special_tokens": ["[food-Negative]", "[food-Positive]", "[food-unknown]", "[food-no majority]", "[ambiance-Negative]", "[ambiance-Positive]", "[ambiance-unknown]", "[ambiance-no majority]", "[noise-Negative]", "[noise-Positive]", "[noise-unknown]", "[noise-no majority]", "[service-Negative]", "[service-Positive]", "[service-unknown]", "[service-no majority]"]}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "roberta-base", "tokenizer_class": "RobertaTokenizer"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1600 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4664005935192108,
|
| 3 |
+
"best_model_checkpoint": "model_output/e2e_opentable_5_way__approximate__0-shot__seed-66__roberta-base/checkpoint-2400",
|
| 4 |
+
"epoch": 4.11663807890223,
|
| 5 |
+
"global_step": 2400,
|
| 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": 15.2337,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.07,
|
| 18 |
+
"learning_rate": 4.956672443674177e-05,
|
| 19 |
+
"loss": 12.7446,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"eval_distillation_accuracy_counterfactual": 0.4099538224730631,
|
| 25 |
+
"eval_distillation_accuracy_factual": 0.8694202154951257,
|
| 26 |
+
"eval_distillation_f1_counterfactual": 0.2767586064145223,
|
| 27 |
+
"eval_distillation_f1_factual": 0.855037145008629,
|
| 28 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4055926115956901,
|
| 29 |
+
"eval_groundtruth_f1_counterfactual": 0.2794447729623582,
|
| 30 |
+
"eval_groundtruth_f1_factual": 0.7140520716279329,
|
| 31 |
+
"eval_icace_cosine": 0.5027520656585693,
|
| 32 |
+
"eval_icace_l2": 0.7894455790519714,
|
| 33 |
+
"eval_icace_normdiff": 0.5822004675865173,
|
| 34 |
+
"eval_loss": 10.286812782287598,
|
| 35 |
+
"eval_runtime": 12.8965,
|
| 36 |
+
"eval_samples_per_second": 302.252,
|
| 37 |
+
"eval_steps_per_second": 2.404,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1,
|
| 42 |
+
"learning_rate": 4.935008665511265e-05,
|
| 43 |
+
"loss": 12.0353,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 4.913344887348354e-05,
|
| 49 |
+
"loss": 10.8775,
|
| 50 |
+
"step": 80
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 4.891681109185442e-05,
|
| 55 |
+
"loss": 10.4023,
|
| 56 |
+
"step": 100
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.17,
|
| 60 |
+
"eval_distillation_accuracy_counterfactual": 0.5100051308363264,
|
| 61 |
+
"eval_distillation_accuracy_factual": 0.8696767573114418,
|
| 62 |
+
"eval_distillation_f1_counterfactual": 0.4037905431000315,
|
| 63 |
+
"eval_distillation_f1_factual": 0.8559902744972321,
|
| 64 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4787070292457671,
|
| 65 |
+
"eval_groundtruth_f1_counterfactual": 0.37533207411941716,
|
| 66 |
+
"eval_groundtruth_f1_factual": 0.7208539604782886,
|
| 67 |
+
"eval_icace_cosine": 0.4815896153450012,
|
| 68 |
+
"eval_icace_l2": 0.6899080872535706,
|
| 69 |
+
"eval_icace_normdiff": 0.4302704632282257,
|
| 70 |
+
"eval_loss": 8.579986572265625,
|
| 71 |
+
"eval_runtime": 12.9731,
|
| 72 |
+
"eval_samples_per_second": 300.469,
|
| 73 |
+
"eval_steps_per_second": 2.39,
|
| 74 |
+
"step": 100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 4.8700173310225307e-05,
|
| 79 |
+
"loss": 10.1421,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24,
|
| 84 |
+
"learning_rate": 4.848353552859619e-05,
|
| 85 |
+
"loss": 9.9736,
|
| 86 |
+
"step": 140
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26,
|
| 90 |
+
"eval_distillation_accuracy_counterfactual": 0.5123140071831709,
|
| 91 |
+
"eval_distillation_accuracy_factual": 0.8655720882503848,
|
| 92 |
+
"eval_distillation_f1_counterfactual": 0.39453533423432713,
|
| 93 |
+
"eval_distillation_f1_factual": 0.8497690853832852,
|
| 94 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5025654181631606,
|
| 95 |
+
"eval_groundtruth_f1_counterfactual": 0.39291773068394925,
|
| 96 |
+
"eval_groundtruth_f1_factual": 0.7016197509963444,
|
| 97 |
+
"eval_icace_cosine": 0.4719853699207306,
|
| 98 |
+
"eval_icace_l2": 0.6965569853782654,
|
| 99 |
+
"eval_icace_normdiff": 0.45266422629356384,
|
| 100 |
+
"eval_loss": 8.405228614807129,
|
| 101 |
+
"eval_runtime": 12.9355,
|
| 102 |
+
"eval_samples_per_second": 301.342,
|
| 103 |
+
"eval_steps_per_second": 2.397,
|
| 104 |
+
"step": 150
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.27,
|
| 108 |
+
"learning_rate": 4.826689774696707e-05,
|
| 109 |
+
"loss": 9.9668,
|
| 110 |
+
"step": 160
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.31,
|
| 114 |
+
"learning_rate": 4.8050259965337955e-05,
|
| 115 |
+
"loss": 9.6499,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.34,
|
| 120 |
+
"learning_rate": 4.7833622183708845e-05,
|
| 121 |
+
"loss": 9.7115,
|
| 122 |
+
"step": 200
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.34,
|
| 126 |
+
"eval_distillation_accuracy_counterfactual": 0.4807593637762955,
|
| 127 |
+
"eval_distillation_accuracy_factual": 0.8599281682914315,
|
| 128 |
+
"eval_distillation_f1_counterfactual": 0.37412255609926637,
|
| 129 |
+
"eval_distillation_f1_factual": 0.8424763100922966,
|
| 130 |
+
"eval_groundtruth_accuracy_counterfactual": 0.45407901487942537,
|
| 131 |
+
"eval_groundtruth_f1_counterfactual": 0.3550761421793228,
|
| 132 |
+
"eval_groundtruth_f1_factual": 0.7263728871661477,
|
| 133 |
+
"eval_icace_cosine": 0.47865259647369385,
|
| 134 |
+
"eval_icace_l2": 0.7145634889602661,
|
| 135 |
+
"eval_icace_normdiff": 0.4607016146183014,
|
| 136 |
+
"eval_loss": 8.778631210327148,
|
| 137 |
+
"eval_runtime": 13.5302,
|
| 138 |
+
"eval_samples_per_second": 288.096,
|
| 139 |
+
"eval_steps_per_second": 2.291,
|
| 140 |
+
"step": 200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.38,
|
| 144 |
+
"learning_rate": 4.761698440207972e-05,
|
| 145 |
+
"loss": 9.6127,
|
| 146 |
+
"step": 220
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.41,
|
| 150 |
+
"learning_rate": 4.740034662045061e-05,
|
| 151 |
+
"loss": 9.643,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.43,
|
| 156 |
+
"eval_distillation_accuracy_counterfactual": 0.4989738327347358,
|
| 157 |
+
"eval_distillation_accuracy_factual": 0.8589020010261673,
|
| 158 |
+
"eval_distillation_f1_counterfactual": 0.4148903629100139,
|
| 159 |
+
"eval_distillation_f1_factual": 0.8451672844026795,
|
| 160 |
+
"eval_groundtruth_accuracy_counterfactual": 0.48460749102103645,
|
| 161 |
+
"eval_groundtruth_f1_counterfactual": 0.4101913073497503,
|
| 162 |
+
"eval_groundtruth_f1_factual": 0.7027318362324826,
|
| 163 |
+
"eval_icace_cosine": 0.4739187955856323,
|
| 164 |
+
"eval_icace_l2": 0.6788182854652405,
|
| 165 |
+
"eval_icace_normdiff": 0.4215588867664337,
|
| 166 |
+
"eval_loss": 8.473577499389648,
|
| 167 |
+
"eval_runtime": 12.9445,
|
| 168 |
+
"eval_samples_per_second": 301.132,
|
| 169 |
+
"eval_steps_per_second": 2.395,
|
| 170 |
+
"step": 250
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45,
|
| 174 |
+
"learning_rate": 4.7183708838821494e-05,
|
| 175 |
+
"loss": 9.4834,
|
| 176 |
+
"step": 260
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.48,
|
| 180 |
+
"learning_rate": 4.6967071057192376e-05,
|
| 181 |
+
"loss": 9.5528,
|
| 182 |
+
"step": 280
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.51,
|
| 186 |
+
"learning_rate": 4.675043327556326e-05,
|
| 187 |
+
"loss": 9.4911,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.51,
|
| 192 |
+
"eval_distillation_accuracy_counterfactual": 0.48024628014366344,
|
| 193 |
+
"eval_distillation_accuracy_factual": 0.8560800410466907,
|
| 194 |
+
"eval_distillation_f1_counterfactual": 0.38759062114466497,
|
| 195 |
+
"eval_distillation_f1_factual": 0.8388069726052654,
|
| 196 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4681888147768086,
|
| 197 |
+
"eval_groundtruth_f1_counterfactual": 0.3863531311941365,
|
| 198 |
+
"eval_groundtruth_f1_factual": 0.7022853957903076,
|
| 199 |
+
"eval_icace_cosine": 0.4793255627155304,
|
| 200 |
+
"eval_icace_l2": 0.6844598054885864,
|
| 201 |
+
"eval_icace_normdiff": 0.41838327050209045,
|
| 202 |
+
"eval_loss": 8.66569709777832,
|
| 203 |
+
"eval_runtime": 12.9451,
|
| 204 |
+
"eval_samples_per_second": 301.117,
|
| 205 |
+
"eval_steps_per_second": 2.395,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.55,
|
| 210 |
+
"learning_rate": 4.653379549393415e-05,
|
| 211 |
+
"loss": 9.2569,
|
| 212 |
+
"step": 320
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.58,
|
| 216 |
+
"learning_rate": 4.6317157712305025e-05,
|
| 217 |
+
"loss": 9.2215,
|
| 218 |
+
"step": 340
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.6,
|
| 222 |
+
"eval_distillation_accuracy_counterfactual": 0.4979476654694715,
|
| 223 |
+
"eval_distillation_accuracy_factual": 0.861723961005644,
|
| 224 |
+
"eval_distillation_f1_counterfactual": 0.4084180941769576,
|
| 225 |
+
"eval_distillation_f1_factual": 0.8476491763825083,
|
| 226 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4976911236531555,
|
| 227 |
+
"eval_groundtruth_f1_counterfactual": 0.41199726943728454,
|
| 228 |
+
"eval_groundtruth_f1_factual": 0.7017647747639904,
|
| 229 |
+
"eval_icace_cosine": 0.48205703496932983,
|
| 230 |
+
"eval_icace_l2": 0.6776481866836548,
|
| 231 |
+
"eval_icace_normdiff": 0.42241370677948,
|
| 232 |
+
"eval_loss": 8.44431209564209,
|
| 233 |
+
"eval_runtime": 14.7727,
|
| 234 |
+
"eval_samples_per_second": 263.866,
|
| 235 |
+
"eval_steps_per_second": 2.098,
|
| 236 |
+
"step": 350
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.62,
|
| 240 |
+
"learning_rate": 4.6100519930675915e-05,
|
| 241 |
+
"loss": 9.0972,
|
| 242 |
+
"step": 360
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.65,
|
| 246 |
+
"learning_rate": 4.58838821490468e-05,
|
| 247 |
+
"loss": 9.1708,
|
| 248 |
+
"step": 380
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.69,
|
| 252 |
+
"learning_rate": 4.566724436741768e-05,
|
| 253 |
+
"loss": 8.8806,
|
| 254 |
+
"step": 400
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.69,
|
| 258 |
+
"eval_distillation_accuracy_counterfactual": 0.5318111852231914,
|
| 259 |
+
"eval_distillation_accuracy_factual": 0.8532580810672139,
|
| 260 |
+
"eval_distillation_f1_counterfactual": 0.45627188162049775,
|
| 261 |
+
"eval_distillation_f1_factual": 0.8433141363137382,
|
| 262 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5146228835300154,
|
| 263 |
+
"eval_groundtruth_f1_counterfactual": 0.4506516610322439,
|
| 264 |
+
"eval_groundtruth_f1_factual": 0.703544919476949,
|
| 265 |
+
"eval_icace_cosine": 0.47363877296447754,
|
| 266 |
+
"eval_icace_l2": 0.6601594090461731,
|
| 267 |
+
"eval_icace_normdiff": 0.4189549684524536,
|
| 268 |
+
"eval_loss": 8.39233684539795,
|
| 269 |
+
"eval_runtime": 13.1803,
|
| 270 |
+
"eval_samples_per_second": 295.745,
|
| 271 |
+
"eval_steps_per_second": 2.352,
|
| 272 |
+
"step": 400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.72,
|
| 276 |
+
"learning_rate": 4.5450606585788563e-05,
|
| 277 |
+
"loss": 9.0014,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.75,
|
| 282 |
+
"learning_rate": 4.5233968804159446e-05,
|
| 283 |
+
"loss": 8.7768,
|
| 284 |
+
"step": 440
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.77,
|
| 288 |
+
"eval_distillation_accuracy_counterfactual": 0.5248845561826577,
|
| 289 |
+
"eval_distillation_accuracy_factual": 0.8471010774756286,
|
| 290 |
+
"eval_distillation_f1_counterfactual": 0.4373197309012829,
|
| 291 |
+
"eval_distillation_f1_factual": 0.8365706913229692,
|
| 292 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5112878399179066,
|
| 293 |
+
"eval_groundtruth_f1_counterfactual": 0.4332310317337549,
|
| 294 |
+
"eval_groundtruth_f1_factual": 0.706420324783281,
|
| 295 |
+
"eval_icace_cosine": 0.47051629424095154,
|
| 296 |
+
"eval_icace_l2": 0.6678978204727173,
|
| 297 |
+
"eval_icace_normdiff": 0.42113974690437317,
|
| 298 |
+
"eval_loss": 8.4581298828125,
|
| 299 |
+
"eval_runtime": 12.9621,
|
| 300 |
+
"eval_samples_per_second": 300.723,
|
| 301 |
+
"eval_steps_per_second": 2.392,
|
| 302 |
+
"step": 450
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.79,
|
| 306 |
+
"learning_rate": 4.501733102253033e-05,
|
| 307 |
+
"loss": 8.8678,
|
| 308 |
+
"step": 460
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.82,
|
| 312 |
+
"learning_rate": 4.480069324090121e-05,
|
| 313 |
+
"loss": 8.9234,
|
| 314 |
+
"step": 480
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.86,
|
| 318 |
+
"learning_rate": 4.45840554592721e-05,
|
| 319 |
+
"loss": 8.8491,
|
| 320 |
+
"step": 500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.86,
|
| 324 |
+
"eval_distillation_accuracy_counterfactual": 0.5269368907131863,
|
| 325 |
+
"eval_distillation_accuracy_factual": 0.8450487429451,
|
| 326 |
+
"eval_distillation_f1_counterfactual": 0.44482702743850444,
|
| 327 |
+
"eval_distillation_f1_factual": 0.8303570404668654,
|
| 328 |
+
"eval_groundtruth_accuracy_counterfactual": 0.517188301693176,
|
| 329 |
+
"eval_groundtruth_f1_counterfactual": 0.44803544206517765,
|
| 330 |
+
"eval_groundtruth_f1_factual": 0.7101076620278356,
|
| 331 |
+
"eval_icace_cosine": 0.4709581732749939,
|
| 332 |
+
"eval_icace_l2": 0.6744429469108582,
|
| 333 |
+
"eval_icace_normdiff": 0.4238233268260956,
|
| 334 |
+
"eval_loss": 8.559771537780762,
|
| 335 |
+
"eval_runtime": 12.9505,
|
| 336 |
+
"eval_samples_per_second": 300.991,
|
| 337 |
+
"eval_steps_per_second": 2.394,
|
| 338 |
+
"step": 500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.89,
|
| 342 |
+
"learning_rate": 4.436741767764298e-05,
|
| 343 |
+
"loss": 8.9139,
|
| 344 |
+
"step": 520
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.93,
|
| 348 |
+
"learning_rate": 4.415077989601387e-05,
|
| 349 |
+
"loss": 8.9491,
|
| 350 |
+
"step": 540
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.94,
|
| 354 |
+
"eval_distillation_accuracy_counterfactual": 0.508722421754746,
|
| 355 |
+
"eval_distillation_accuracy_factual": 0.8270908158029758,
|
| 356 |
+
"eval_distillation_f1_counterfactual": 0.4457012957498576,
|
| 357 |
+
"eval_distillation_f1_factual": 0.8156237227220284,
|
| 358 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4987172909184197,
|
| 359 |
+
"eval_groundtruth_f1_counterfactual": 0.4444355597987831,
|
| 360 |
+
"eval_groundtruth_f1_factual": 0.704491382874623,
|
| 361 |
+
"eval_icace_cosine": 0.4708389639854431,
|
| 362 |
+
"eval_icace_l2": 0.6740121245384216,
|
| 363 |
+
"eval_icace_normdiff": 0.4291449189186096,
|
| 364 |
+
"eval_loss": 8.710935592651367,
|
| 365 |
+
"eval_runtime": 13.1041,
|
| 366 |
+
"eval_samples_per_second": 297.465,
|
| 367 |
+
"eval_steps_per_second": 2.366,
|
| 368 |
+
"step": 550
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.96,
|
| 372 |
+
"learning_rate": 4.393414211438475e-05,
|
| 373 |
+
"loss": 8.6501,
|
| 374 |
+
"step": 560
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.99,
|
| 378 |
+
"learning_rate": 4.371750433275563e-05,
|
| 379 |
+
"loss": 8.8115,
|
| 380 |
+
"step": 580
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.03,
|
| 384 |
+
"learning_rate": 4.3500866551126516e-05,
|
| 385 |
+
"loss": 8.5726,
|
| 386 |
+
"step": 600
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.03,
|
| 390 |
+
"eval_distillation_accuracy_counterfactual": 0.5061570035915854,
|
| 391 |
+
"eval_distillation_accuracy_factual": 0.8486403283735249,
|
| 392 |
+
"eval_distillation_f1_counterfactual": 0.41610676926533285,
|
| 393 |
+
"eval_distillation_f1_factual": 0.8310444475559301,
|
| 394 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4905079527963058,
|
| 395 |
+
"eval_groundtruth_f1_counterfactual": 0.4145735924063139,
|
| 396 |
+
"eval_groundtruth_f1_factual": 0.7092515264551655,
|
| 397 |
+
"eval_icace_cosine": 0.4776078760623932,
|
| 398 |
+
"eval_icace_l2": 0.6925671100616455,
|
| 399 |
+
"eval_icace_normdiff": 0.42703282833099365,
|
| 400 |
+
"eval_loss": 8.773487091064453,
|
| 401 |
+
"eval_runtime": 13.0705,
|
| 402 |
+
"eval_samples_per_second": 298.228,
|
| 403 |
+
"eval_steps_per_second": 2.372,
|
| 404 |
+
"step": 600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.06,
|
| 408 |
+
"learning_rate": 4.3284228769497406e-05,
|
| 409 |
+
"loss": 8.7449,
|
| 410 |
+
"step": 620
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.1,
|
| 414 |
+
"learning_rate": 4.306759098786828e-05,
|
| 415 |
+
"loss": 8.5654,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.11,
|
| 420 |
+
"eval_distillation_accuracy_counterfactual": 0.5148794253463315,
|
| 421 |
+
"eval_distillation_accuracy_factual": 0.8314520266803489,
|
| 422 |
+
"eval_distillation_f1_counterfactual": 0.4350161586459501,
|
| 423 |
+
"eval_distillation_f1_factual": 0.8207900808818138,
|
| 424 |
+
"eval_groundtruth_accuracy_counterfactual": 0.513340174448435,
|
| 425 |
+
"eval_groundtruth_f1_counterfactual": 0.4424165935664268,
|
| 426 |
+
"eval_groundtruth_f1_factual": 0.6858850991781995,
|
| 427 |
+
"eval_icace_cosine": 0.4748988747596741,
|
| 428 |
+
"eval_icace_l2": 0.6583500504493713,
|
| 429 |
+
"eval_icace_normdiff": 0.4023895561695099,
|
| 430 |
+
"eval_loss": 8.626461029052734,
|
| 431 |
+
"eval_runtime": 13.1109,
|
| 432 |
+
"eval_samples_per_second": 297.31,
|
| 433 |
+
"eval_steps_per_second": 2.364,
|
| 434 |
+
"step": 650
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.13,
|
| 438 |
+
"learning_rate": 4.285095320623917e-05,
|
| 439 |
+
"loss": 8.5985,
|
| 440 |
+
"step": 660
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.17,
|
| 444 |
+
"learning_rate": 4.2634315424610055e-05,
|
| 445 |
+
"loss": 8.9032,
|
| 446 |
+
"step": 680
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.2,
|
| 450 |
+
"learning_rate": 4.241767764298094e-05,
|
| 451 |
+
"loss": 8.5234,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.2,
|
| 456 |
+
"eval_distillation_accuracy_counterfactual": 0.5266803488968702,
|
| 457 |
+
"eval_distillation_accuracy_factual": 0.8347870702924577,
|
| 458 |
+
"eval_distillation_f1_counterfactual": 0.44418104449377827,
|
| 459 |
+
"eval_distillation_f1_factual": 0.8166622315451695,
|
| 460 |
+
"eval_groundtruth_accuracy_counterfactual": 0.512570548999487,
|
| 461 |
+
"eval_groundtruth_f1_counterfactual": 0.4399064148597627,
|
| 462 |
+
"eval_groundtruth_f1_factual": 0.7003621420252235,
|
| 463 |
+
"eval_icace_cosine": 0.4755347669124603,
|
| 464 |
+
"eval_icace_l2": 0.6643328666687012,
|
| 465 |
+
"eval_icace_normdiff": 0.4135916829109192,
|
| 466 |
+
"eval_loss": 8.528804779052734,
|
| 467 |
+
"eval_runtime": 12.9782,
|
| 468 |
+
"eval_samples_per_second": 300.35,
|
| 469 |
+
"eval_steps_per_second": 2.389,
|
| 470 |
+
"step": 700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.23,
|
| 474 |
+
"learning_rate": 4.220103986135182e-05,
|
| 475 |
+
"loss": 8.3629,
|
| 476 |
+
"step": 720
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.27,
|
| 480 |
+
"learning_rate": 4.198440207972271e-05,
|
| 481 |
+
"loss": 8.6834,
|
| 482 |
+
"step": 740
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.29,
|
| 486 |
+
"eval_distillation_accuracy_counterfactual": 0.5253976398152899,
|
| 487 |
+
"eval_distillation_accuracy_factual": 0.8499230374551052,
|
| 488 |
+
"eval_distillation_f1_counterfactual": 0.4481280211692343,
|
| 489 |
+
"eval_distillation_f1_factual": 0.8427632254580978,
|
| 490 |
+
"eval_groundtruth_accuracy_counterfactual": 0.521292970754233,
|
| 491 |
+
"eval_groundtruth_f1_counterfactual": 0.4535188471820635,
|
| 492 |
+
"eval_groundtruth_f1_factual": 0.7050025057202923,
|
| 493 |
+
"eval_icace_cosine": 0.478320837020874,
|
| 494 |
+
"eval_icace_l2": 0.6507540345191956,
|
| 495 |
+
"eval_icace_normdiff": 0.4042801558971405,
|
| 496 |
+
"eval_loss": 8.547685623168945,
|
| 497 |
+
"eval_runtime": 13.0628,
|
| 498 |
+
"eval_samples_per_second": 298.406,
|
| 499 |
+
"eval_steps_per_second": 2.373,
|
| 500 |
+
"step": 750
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3,
|
| 504 |
+
"learning_rate": 4.1767764298093586e-05,
|
| 505 |
+
"loss": 8.373,
|
| 506 |
+
"step": 760
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.34,
|
| 510 |
+
"learning_rate": 4.1551126516464476e-05,
|
| 511 |
+
"loss": 8.545,
|
| 512 |
+
"step": 780
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.37,
|
| 516 |
+
"learning_rate": 4.133448873483536e-05,
|
| 517 |
+
"loss": 8.3301,
|
| 518 |
+
"step": 800
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.37,
|
| 522 |
+
"eval_distillation_accuracy_counterfactual": 0.5246280143663418,
|
| 523 |
+
"eval_distillation_accuracy_factual": 0.857875833760903,
|
| 524 |
+
"eval_distillation_f1_counterfactual": 0.43678271154246406,
|
| 525 |
+
"eval_distillation_f1_factual": 0.8404426089149801,
|
| 526 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5177013853258081,
|
| 527 |
+
"eval_groundtruth_f1_counterfactual": 0.4372037073522039,
|
| 528 |
+
"eval_groundtruth_f1_factual": 0.7164819803959707,
|
| 529 |
+
"eval_icace_cosine": 0.4767405092716217,
|
| 530 |
+
"eval_icace_l2": 0.6505340933799744,
|
| 531 |
+
"eval_icace_normdiff": 0.3892482817173004,
|
| 532 |
+
"eval_loss": 8.437895774841309,
|
| 533 |
+
"eval_runtime": 12.9829,
|
| 534 |
+
"eval_samples_per_second": 300.242,
|
| 535 |
+
"eval_steps_per_second": 2.388,
|
| 536 |
+
"step": 800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.41,
|
| 540 |
+
"learning_rate": 4.111785095320624e-05,
|
| 541 |
+
"loss": 8.5471,
|
| 542 |
+
"step": 820
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.44,
|
| 546 |
+
"learning_rate": 4.0901213171577124e-05,
|
| 547 |
+
"loss": 8.5899,
|
| 548 |
+
"step": 840
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.46,
|
| 552 |
+
"eval_distillation_accuracy_counterfactual": 0.5389943560800411,
|
| 553 |
+
"eval_distillation_accuracy_factual": 0.8668547973319651,
|
| 554 |
+
"eval_distillation_f1_counterfactual": 0.43508558636235506,
|
| 555 |
+
"eval_distillation_f1_factual": 0.8544181051951583,
|
| 556 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5287326834273987,
|
| 557 |
+
"eval_groundtruth_f1_counterfactual": 0.4313681541912551,
|
| 558 |
+
"eval_groundtruth_f1_factual": 0.7021443551588085,
|
| 559 |
+
"eval_icace_cosine": 0.47022899985313416,
|
| 560 |
+
"eval_icace_l2": 0.6471868753433228,
|
| 561 |
+
"eval_icace_normdiff": 0.3974311053752899,
|
| 562 |
+
"eval_loss": 8.187715530395508,
|
| 563 |
+
"eval_runtime": 12.9448,
|
| 564 |
+
"eval_samples_per_second": 301.125,
|
| 565 |
+
"eval_steps_per_second": 2.395,
|
| 566 |
+
"step": 850
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.48,
|
| 570 |
+
"learning_rate": 4.068457538994801e-05,
|
| 571 |
+
"loss": 8.4412,
|
| 572 |
+
"step": 860
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.51,
|
| 576 |
+
"learning_rate": 4.04679376083189e-05,
|
| 577 |
+
"loss": 8.5992,
|
| 578 |
+
"step": 880
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.54,
|
| 582 |
+
"learning_rate": 4.025129982668977e-05,
|
| 583 |
+
"loss": 8.5057,
|
| 584 |
+
"step": 900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.54,
|
| 588 |
+
"eval_distillation_accuracy_counterfactual": 0.5338635197537198,
|
| 589 |
+
"eval_distillation_accuracy_factual": 0.8494099538224731,
|
| 590 |
+
"eval_distillation_f1_counterfactual": 0.4538940008145664,
|
| 591 |
+
"eval_distillation_f1_factual": 0.8393372080212285,
|
| 592 |
+
"eval_groundtruth_accuracy_counterfactual": 0.521549512570549,
|
| 593 |
+
"eval_groundtruth_f1_counterfactual": 0.44602586407123895,
|
| 594 |
+
"eval_groundtruth_f1_factual": 0.7062665719976977,
|
| 595 |
+
"eval_icace_cosine": 0.47390881180763245,
|
| 596 |
+
"eval_icace_l2": 0.6635516881942749,
|
| 597 |
+
"eval_icace_normdiff": 0.41677233576774597,
|
| 598 |
+
"eval_loss": 8.466891288757324,
|
| 599 |
+
"eval_runtime": 12.9586,
|
| 600 |
+
"eval_samples_per_second": 300.803,
|
| 601 |
+
"eval_steps_per_second": 2.392,
|
| 602 |
+
"step": 900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.58,
|
| 606 |
+
"learning_rate": 4.003466204506066e-05,
|
| 607 |
+
"loss": 8.5256,
|
| 608 |
+
"step": 920
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.61,
|
| 612 |
+
"learning_rate": 3.9818024263431546e-05,
|
| 613 |
+
"loss": 8.4663,
|
| 614 |
+
"step": 940
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.63,
|
| 618 |
+
"eval_distillation_accuracy_counterfactual": 0.539250897896357,
|
| 619 |
+
"eval_distillation_accuracy_factual": 0.8429964084145716,
|
| 620 |
+
"eval_distillation_f1_counterfactual": 0.4552348813290624,
|
| 621 |
+
"eval_distillation_f1_factual": 0.8279955804250164,
|
| 622 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5248845561826577,
|
| 623 |
+
"eval_groundtruth_f1_counterfactual": 0.44541029004084864,
|
| 624 |
+
"eval_groundtruth_f1_factual": 0.7174847360382134,
|
| 625 |
+
"eval_icace_cosine": 0.47177237272262573,
|
| 626 |
+
"eval_icace_l2": 0.6441071629524231,
|
| 627 |
+
"eval_icace_normdiff": 0.3987690806388855,
|
| 628 |
+
"eval_loss": 8.24371337890625,
|
| 629 |
+
"eval_runtime": 13.1576,
|
| 630 |
+
"eval_samples_per_second": 296.255,
|
| 631 |
+
"eval_steps_per_second": 2.356,
|
| 632 |
+
"step": 950
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.65,
|
| 636 |
+
"learning_rate": 3.960138648180243e-05,
|
| 637 |
+
"loss": 8.4244,
|
| 638 |
+
"step": 960
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.68,
|
| 642 |
+
"learning_rate": 3.938474870017331e-05,
|
| 643 |
+
"loss": 8.5347,
|
| 644 |
+
"step": 980
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.72,
|
| 648 |
+
"learning_rate": 3.91681109185442e-05,
|
| 649 |
+
"loss": 8.5181,
|
| 650 |
+
"step": 1000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.72,
|
| 654 |
+
"eval_distillation_accuracy_counterfactual": 0.5523345305284761,
|
| 655 |
+
"eval_distillation_accuracy_factual": 0.8568496664956388,
|
| 656 |
+
"eval_distillation_f1_counterfactual": 0.43192107651721756,
|
| 657 |
+
"eval_distillation_f1_factual": 0.8446400408995643,
|
| 658 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5310415597742432,
|
| 659 |
+
"eval_groundtruth_f1_counterfactual": 0.4221211937230137,
|
| 660 |
+
"eval_groundtruth_f1_factual": 0.7176363988738272,
|
| 661 |
+
"eval_icace_cosine": 0.4691796898841858,
|
| 662 |
+
"eval_icace_l2": 0.6469516158103943,
|
| 663 |
+
"eval_icace_normdiff": 0.40605321526527405,
|
| 664 |
+
"eval_loss": 8.24507999420166,
|
| 665 |
+
"eval_runtime": 12.9758,
|
| 666 |
+
"eval_samples_per_second": 300.405,
|
| 667 |
+
"eval_steps_per_second": 2.389,
|
| 668 |
+
"step": 1000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.75,
|
| 672 |
+
"learning_rate": 3.895147313691508e-05,
|
| 673 |
+
"loss": 8.4384,
|
| 674 |
+
"step": 1020
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.78,
|
| 678 |
+
"learning_rate": 3.873483535528597e-05,
|
| 679 |
+
"loss": 8.4464,
|
| 680 |
+
"step": 1040
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.8,
|
| 684 |
+
"eval_distillation_accuracy_counterfactual": 0.5341200615700359,
|
| 685 |
+
"eval_distillation_accuracy_factual": 0.8471010774756286,
|
| 686 |
+
"eval_distillation_f1_counterfactual": 0.44797373858638706,
|
| 687 |
+
"eval_distillation_f1_factual": 0.8352221547782517,
|
| 688 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5187275525910724,
|
| 689 |
+
"eval_groundtruth_f1_counterfactual": 0.438959174201451,
|
| 690 |
+
"eval_groundtruth_f1_factual": 0.7045395682269071,
|
| 691 |
+
"eval_icace_cosine": 0.4699629247188568,
|
| 692 |
+
"eval_icace_l2": 0.650560200214386,
|
| 693 |
+
"eval_icace_normdiff": 0.4021012485027313,
|
| 694 |
+
"eval_loss": 8.46190357208252,
|
| 695 |
+
"eval_runtime": 12.9778,
|
| 696 |
+
"eval_samples_per_second": 300.359,
|
| 697 |
+
"eval_steps_per_second": 2.389,
|
| 698 |
+
"step": 1050
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.82,
|
| 702 |
+
"learning_rate": 3.851819757365685e-05,
|
| 703 |
+
"loss": 8.3977,
|
| 704 |
+
"step": 1060
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.85,
|
| 708 |
+
"learning_rate": 3.830155979202773e-05,
|
| 709 |
+
"loss": 8.4469,
|
| 710 |
+
"step": 1080
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.89,
|
| 714 |
+
"learning_rate": 3.8084922010398616e-05,
|
| 715 |
+
"loss": 8.4589,
|
| 716 |
+
"step": 1100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.89,
|
| 720 |
+
"eval_distillation_accuracy_counterfactual": 0.538481272447409,
|
| 721 |
+
"eval_distillation_accuracy_factual": 0.853771164699846,
|
| 722 |
+
"eval_distillation_f1_counterfactual": 0.4574351380443254,
|
| 723 |
+
"eval_distillation_f1_factual": 0.8413133626862945,
|
| 724 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5338635197537198,
|
| 725 |
+
"eval_groundtruth_f1_counterfactual": 0.4618955956329911,
|
| 726 |
+
"eval_groundtruth_f1_factual": 0.712001520958305,
|
| 727 |
+
"eval_icace_cosine": 0.47186800837516785,
|
| 728 |
+
"eval_icace_l2": 0.6550236940383911,
|
| 729 |
+
"eval_icace_normdiff": 0.4062390625476837,
|
| 730 |
+
"eval_loss": 8.270210266113281,
|
| 731 |
+
"eval_runtime": 13.0076,
|
| 732 |
+
"eval_samples_per_second": 299.671,
|
| 733 |
+
"eval_steps_per_second": 2.383,
|
| 734 |
+
"step": 1100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.92,
|
| 738 |
+
"learning_rate": 3.78682842287695e-05,
|
| 739 |
+
"loss": 8.2726,
|
| 740 |
+
"step": 1120
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.96,
|
| 744 |
+
"learning_rate": 3.765164644714038e-05,
|
| 745 |
+
"loss": 8.543,
|
| 746 |
+
"step": 1140
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.97,
|
| 750 |
+
"eval_distillation_accuracy_counterfactual": 0.5454079014879425,
|
| 751 |
+
"eval_distillation_accuracy_factual": 0.8565931246793227,
|
| 752 |
+
"eval_distillation_f1_counterfactual": 0.4358856901648176,
|
| 753 |
+
"eval_distillation_f1_factual": 0.8450511347625259,
|
| 754 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5369420215495125,
|
| 755 |
+
"eval_groundtruth_f1_counterfactual": 0.4328763616035284,
|
| 756 |
+
"eval_groundtruth_f1_factual": 0.7122842482342289,
|
| 757 |
+
"eval_icace_cosine": 0.470956027507782,
|
| 758 |
+
"eval_icace_l2": 0.6414260864257812,
|
| 759 |
+
"eval_icace_normdiff": 0.3959381878376007,
|
| 760 |
+
"eval_loss": 8.278718948364258,
|
| 761 |
+
"eval_runtime": 13.0747,
|
| 762 |
+
"eval_samples_per_second": 298.133,
|
| 763 |
+
"eval_steps_per_second": 2.371,
|
| 764 |
+
"step": 1150
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.99,
|
| 768 |
+
"learning_rate": 3.7435008665511264e-05,
|
| 769 |
+
"loss": 8.5847,
|
| 770 |
+
"step": 1160
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.02,
|
| 774 |
+
"learning_rate": 3.7218370883882154e-05,
|
| 775 |
+
"loss": 8.495,
|
| 776 |
+
"step": 1180
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.06,
|
| 780 |
+
"learning_rate": 3.700173310225303e-05,
|
| 781 |
+
"loss": 8.2151,
|
| 782 |
+
"step": 1200
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.06,
|
| 786 |
+
"eval_distillation_accuracy_counterfactual": 0.5323242688558235,
|
| 787 |
+
"eval_distillation_accuracy_factual": 0.844792201128784,
|
| 788 |
+
"eval_distillation_f1_counterfactual": 0.4578721834749787,
|
| 789 |
+
"eval_distillation_f1_factual": 0.8335573328949983,
|
| 790 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5205233453052848,
|
| 791 |
+
"eval_groundtruth_f1_counterfactual": 0.4529105959701748,
|
| 792 |
+
"eval_groundtruth_f1_factual": 0.6982997202664809,
|
| 793 |
+
"eval_icace_cosine": 0.4770417809486389,
|
| 794 |
+
"eval_icace_l2": 0.664779782295227,
|
| 795 |
+
"eval_icace_normdiff": 0.4089834690093994,
|
| 796 |
+
"eval_loss": 8.583990097045898,
|
| 797 |
+
"eval_runtime": 12.9637,
|
| 798 |
+
"eval_samples_per_second": 300.686,
|
| 799 |
+
"eval_steps_per_second": 2.391,
|
| 800 |
+
"step": 1200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.09,
|
| 804 |
+
"learning_rate": 3.678509532062392e-05,
|
| 805 |
+
"loss": 8.2576,
|
| 806 |
+
"step": 1220
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.13,
|
| 810 |
+
"learning_rate": 3.65684575389948e-05,
|
| 811 |
+
"loss": 8.3456,
|
| 812 |
+
"step": 1240
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.14,
|
| 816 |
+
"eval_distillation_accuracy_counterfactual": 0.5210364289379169,
|
| 817 |
+
"eval_distillation_accuracy_factual": 0.8414571575166753,
|
| 818 |
+
"eval_distillation_f1_counterfactual": 0.44541850110843295,
|
| 819 |
+
"eval_distillation_f1_factual": 0.8216034713991617,
|
| 820 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5228322216521293,
|
| 821 |
+
"eval_groundtruth_f1_counterfactual": 0.4520848884145957,
|
| 822 |
+
"eval_groundtruth_f1_factual": 0.7136759478326009,
|
| 823 |
+
"eval_icace_cosine": 0.48024165630340576,
|
| 824 |
+
"eval_icace_l2": 0.6682620048522949,
|
| 825 |
+
"eval_icace_normdiff": 0.41649964451789856,
|
| 826 |
+
"eval_loss": 8.683144569396973,
|
| 827 |
+
"eval_runtime": 12.9156,
|
| 828 |
+
"eval_samples_per_second": 301.806,
|
| 829 |
+
"eval_steps_per_second": 2.4,
|
| 830 |
+
"step": 1250
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.16,
|
| 834 |
+
"learning_rate": 3.6351819757365686e-05,
|
| 835 |
+
"loss": 8.2687,
|
| 836 |
+
"step": 1260
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2,
|
| 840 |
+
"learning_rate": 3.613518197573657e-05,
|
| 841 |
+
"loss": 8.343,
|
| 842 |
+
"step": 1280
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.23,
|
| 846 |
+
"learning_rate": 3.591854419410746e-05,
|
| 847 |
+
"loss": 8.2615,
|
| 848 |
+
"step": 1300
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.23,
|
| 852 |
+
"eval_distillation_accuracy_counterfactual": 0.5210364289379169,
|
| 853 |
+
"eval_distillation_accuracy_factual": 0.8376090302719343,
|
| 854 |
+
"eval_distillation_f1_counterfactual": 0.4470542011154386,
|
| 855 |
+
"eval_distillation_f1_factual": 0.8243930760094701,
|
| 856 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5094920472036942,
|
| 857 |
+
"eval_groundtruth_f1_counterfactual": 0.4424943238928763,
|
| 858 |
+
"eval_groundtruth_f1_factual": 0.7180425551041923,
|
| 859 |
+
"eval_icace_cosine": 0.4772956669330597,
|
| 860 |
+
"eval_icace_l2": 0.675339937210083,
|
| 861 |
+
"eval_icace_normdiff": 0.4207656979560852,
|
| 862 |
+
"eval_loss": 8.777983665466309,
|
| 863 |
+
"eval_runtime": 12.95,
|
| 864 |
+
"eval_samples_per_second": 301.004,
|
| 865 |
+
"eval_steps_per_second": 2.394,
|
| 866 |
+
"step": 1300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.26,
|
| 870 |
+
"learning_rate": 3.5701906412478334e-05,
|
| 871 |
+
"loss": 8.3164,
|
| 872 |
+
"step": 1320
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.3,
|
| 876 |
+
"learning_rate": 3.5485268630849224e-05,
|
| 877 |
+
"loss": 8.2886,
|
| 878 |
+
"step": 1340
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.32,
|
| 882 |
+
"eval_distillation_accuracy_counterfactual": 0.5233453052847614,
|
| 883 |
+
"eval_distillation_accuracy_factual": 0.8563365828630066,
|
| 884 |
+
"eval_distillation_f1_counterfactual": 0.4352218638479065,
|
| 885 |
+
"eval_distillation_f1_factual": 0.842128331673918,
|
| 886 |
+
"eval_groundtruth_accuracy_counterfactual": 0.504617752693689,
|
| 887 |
+
"eval_groundtruth_f1_counterfactual": 0.42672300871162205,
|
| 888 |
+
"eval_groundtruth_f1_factual": 0.7038431621141455,
|
| 889 |
+
"eval_icace_cosine": 0.47232314944267273,
|
| 890 |
+
"eval_icace_l2": 0.6612510681152344,
|
| 891 |
+
"eval_icace_normdiff": 0.4119323790073395,
|
| 892 |
+
"eval_loss": 8.546701431274414,
|
| 893 |
+
"eval_runtime": 12.8815,
|
| 894 |
+
"eval_samples_per_second": 302.605,
|
| 895 |
+
"eval_steps_per_second": 2.407,
|
| 896 |
+
"step": 1350
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.33,
|
| 900 |
+
"learning_rate": 3.526863084922011e-05,
|
| 901 |
+
"loss": 8.0323,
|
| 902 |
+
"step": 1360
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.37,
|
| 906 |
+
"learning_rate": 3.505199306759099e-05,
|
| 907 |
+
"loss": 8.2412,
|
| 908 |
+
"step": 1380
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.4,
|
| 912 |
+
"learning_rate": 3.483535528596187e-05,
|
| 913 |
+
"loss": 8.2755,
|
| 914 |
+
"step": 1400
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.4,
|
| 918 |
+
"eval_distillation_accuracy_counterfactual": 0.5279630579784504,
|
| 919 |
+
"eval_distillation_accuracy_factual": 0.8512057465366855,
|
| 920 |
+
"eval_distillation_f1_counterfactual": 0.42885545419249443,
|
| 921 |
+
"eval_distillation_f1_factual": 0.8336172051549792,
|
| 922 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5110312981015905,
|
| 923 |
+
"eval_groundtruth_f1_counterfactual": 0.4182726548657759,
|
| 924 |
+
"eval_groundtruth_f1_factual": 0.7081817865697444,
|
| 925 |
+
"eval_icace_cosine": 0.47038188576698303,
|
| 926 |
+
"eval_icace_l2": 0.6541587114334106,
|
| 927 |
+
"eval_icace_normdiff": 0.4046129286289215,
|
| 928 |
+
"eval_loss": 8.557218551635742,
|
| 929 |
+
"eval_runtime": 12.9809,
|
| 930 |
+
"eval_samples_per_second": 300.286,
|
| 931 |
+
"eval_steps_per_second": 2.388,
|
| 932 |
+
"step": 1400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.44,
|
| 936 |
+
"learning_rate": 3.461871750433276e-05,
|
| 937 |
+
"loss": 8.1159,
|
| 938 |
+
"step": 1420
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.47,
|
| 942 |
+
"learning_rate": 3.440207972270364e-05,
|
| 943 |
+
"loss": 8.2512,
|
| 944 |
+
"step": 1440
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.49,
|
| 948 |
+
"eval_distillation_accuracy_counterfactual": 0.5048742945100051,
|
| 949 |
+
"eval_distillation_accuracy_factual": 0.8442791174961519,
|
| 950 |
+
"eval_distillation_f1_counterfactual": 0.4471986183427893,
|
| 951 |
+
"eval_distillation_f1_factual": 0.8328539372383206,
|
| 952 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5061570035915854,
|
| 953 |
+
"eval_groundtruth_f1_counterfactual": 0.45459231049993376,
|
| 954 |
+
"eval_groundtruth_f1_factual": 0.7243435837992515,
|
| 955 |
+
"eval_icace_cosine": 0.47215959429740906,
|
| 956 |
+
"eval_icace_l2": 0.6676862239837646,
|
| 957 |
+
"eval_icace_normdiff": 0.4189312160015106,
|
| 958 |
+
"eval_loss": 8.632984161376953,
|
| 959 |
+
"eval_runtime": 12.9685,
|
| 960 |
+
"eval_samples_per_second": 300.574,
|
| 961 |
+
"eval_steps_per_second": 2.39,
|
| 962 |
+
"step": 1450
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.5,
|
| 966 |
+
"learning_rate": 3.418544194107453e-05,
|
| 967 |
+
"loss": 8.4124,
|
| 968 |
+
"step": 1460
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.54,
|
| 972 |
+
"learning_rate": 3.396880415944541e-05,
|
| 973 |
+
"loss": 8.2396,
|
| 974 |
+
"step": 1480
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.57,
|
| 978 |
+
"learning_rate": 3.3752166377816294e-05,
|
| 979 |
+
"loss": 8.1802,
|
| 980 |
+
"step": 1500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.57,
|
| 984 |
+
"eval_distillation_accuracy_counterfactual": 0.5266803488968702,
|
| 985 |
+
"eval_distillation_accuracy_factual": 0.8424833247819394,
|
| 986 |
+
"eval_distillation_f1_counterfactual": 0.4296539407490461,
|
| 987 |
+
"eval_distillation_f1_factual": 0.8232924538803463,
|
| 988 |
+
"eval_groundtruth_accuracy_counterfactual": 0.521292970754233,
|
| 989 |
+
"eval_groundtruth_f1_counterfactual": 0.4329058498324728,
|
| 990 |
+
"eval_groundtruth_f1_factual": 0.693798726302169,
|
| 991 |
+
"eval_icace_cosine": 0.47156795859336853,
|
| 992 |
+
"eval_icace_l2": 0.6467342972755432,
|
| 993 |
+
"eval_icace_normdiff": 0.39208564162254333,
|
| 994 |
+
"eval_loss": 8.581865310668945,
|
| 995 |
+
"eval_runtime": 12.9632,
|
| 996 |
+
"eval_samples_per_second": 300.697,
|
| 997 |
+
"eval_steps_per_second": 2.391,
|
| 998 |
+
"step": 1500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.61,
|
| 1002 |
+
"learning_rate": 3.353552859618718e-05,
|
| 1003 |
+
"loss": 8.0483,
|
| 1004 |
+
"step": 1520
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.64,
|
| 1008 |
+
"learning_rate": 3.331889081455806e-05,
|
| 1009 |
+
"loss": 8.3776,
|
| 1010 |
+
"step": 1540
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.66,
|
| 1014 |
+
"eval_distillation_accuracy_counterfactual": 0.5395074397126731,
|
| 1015 |
+
"eval_distillation_accuracy_factual": 0.8624935864545921,
|
| 1016 |
+
"eval_distillation_f1_counterfactual": 0.4487826083438827,
|
| 1017 |
+
"eval_distillation_f1_factual": 0.8447571530576526,
|
| 1018 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5233453052847614,
|
| 1019 |
+
"eval_groundtruth_f1_counterfactual": 0.44381794658084023,
|
| 1020 |
+
"eval_groundtruth_f1_factual": 0.7228365951086511,
|
| 1021 |
+
"eval_icace_cosine": 0.467881977558136,
|
| 1022 |
+
"eval_icace_l2": 0.6500933766365051,
|
| 1023 |
+
"eval_icace_normdiff": 0.4116222858428955,
|
| 1024 |
+
"eval_loss": 8.251863479614258,
|
| 1025 |
+
"eval_runtime": 13.0173,
|
| 1026 |
+
"eval_samples_per_second": 299.447,
|
| 1027 |
+
"eval_steps_per_second": 2.381,
|
| 1028 |
+
"step": 1550
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.68,
|
| 1032 |
+
"learning_rate": 3.310225303292894e-05,
|
| 1033 |
+
"loss": 8.3191,
|
| 1034 |
+
"step": 1560
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.71,
|
| 1038 |
+
"learning_rate": 3.2885615251299825e-05,
|
| 1039 |
+
"loss": 8.3499,
|
| 1040 |
+
"step": 1580
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.74,
|
| 1044 |
+
"learning_rate": 3.2668977469670715e-05,
|
| 1045 |
+
"loss": 8.2759,
|
| 1046 |
+
"step": 1600
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.74,
|
| 1050 |
+
"eval_distillation_accuracy_counterfactual": 0.5264238070805541,
|
| 1051 |
+
"eval_distillation_accuracy_factual": 0.8381221139045665,
|
| 1052 |
+
"eval_distillation_f1_counterfactual": 0.4640025505726335,
|
| 1053 |
+
"eval_distillation_f1_factual": 0.8226839825314822,
|
| 1054 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5112878399179066,
|
| 1055 |
+
"eval_groundtruth_f1_counterfactual": 0.4572176402176623,
|
| 1056 |
+
"eval_groundtruth_f1_factual": 0.7122677529021709,
|
| 1057 |
+
"eval_icace_cosine": 0.46834951639175415,
|
| 1058 |
+
"eval_icace_l2": 0.6554086208343506,
|
| 1059 |
+
"eval_icace_normdiff": 0.4115924835205078,
|
| 1060 |
+
"eval_loss": 8.593127250671387,
|
| 1061 |
+
"eval_runtime": 12.9442,
|
| 1062 |
+
"eval_samples_per_second": 301.138,
|
| 1063 |
+
"eval_steps_per_second": 2.395,
|
| 1064 |
+
"step": 1600
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.78,
|
| 1068 |
+
"learning_rate": 3.245233968804159e-05,
|
| 1069 |
+
"loss": 8.2134,
|
| 1070 |
+
"step": 1620
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.81,
|
| 1074 |
+
"learning_rate": 3.223570190641248e-05,
|
| 1075 |
+
"loss": 8.2762,
|
| 1076 |
+
"step": 1640
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.83,
|
| 1080 |
+
"eval_distillation_accuracy_counterfactual": 0.5274499743458184,
|
| 1081 |
+
"eval_distillation_accuracy_factual": 0.8532580810672139,
|
| 1082 |
+
"eval_distillation_f1_counterfactual": 0.4582506359486283,
|
| 1083 |
+
"eval_distillation_f1_factual": 0.8369130644925307,
|
| 1084 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5105182144689584,
|
| 1085 |
+
"eval_groundtruth_f1_counterfactual": 0.4492512316522279,
|
| 1086 |
+
"eval_groundtruth_f1_factual": 0.7137344968653884,
|
| 1087 |
+
"eval_icace_cosine": 0.4689863324165344,
|
| 1088 |
+
"eval_icace_l2": 0.6501835584640503,
|
| 1089 |
+
"eval_icace_normdiff": 0.40152183175086975,
|
| 1090 |
+
"eval_loss": 8.484435081481934,
|
| 1091 |
+
"eval_runtime": 12.9458,
|
| 1092 |
+
"eval_samples_per_second": 301.102,
|
| 1093 |
+
"eval_steps_per_second": 2.395,
|
| 1094 |
+
"step": 1650
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.85,
|
| 1098 |
+
"learning_rate": 3.2019064124783364e-05,
|
| 1099 |
+
"loss": 8.3236,
|
| 1100 |
+
"step": 1660
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.88,
|
| 1104 |
+
"learning_rate": 3.1802426343154247e-05,
|
| 1105 |
+
"loss": 8.1251,
|
| 1106 |
+
"step": 1680
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.92,
|
| 1110 |
+
"learning_rate": 3.158578856152513e-05,
|
| 1111 |
+
"loss": 8.3658,
|
| 1112 |
+
"step": 1700
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.92,
|
| 1116 |
+
"eval_distillation_accuracy_counterfactual": 0.5236018471010775,
|
| 1117 |
+
"eval_distillation_accuracy_factual": 0.8442791174961519,
|
| 1118 |
+
"eval_distillation_f1_counterfactual": 0.4500677919517996,
|
| 1119 |
+
"eval_distillation_f1_factual": 0.8294190656695536,
|
| 1120 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5161621344279117,
|
| 1121 |
+
"eval_groundtruth_f1_counterfactual": 0.44921994678493393,
|
| 1122 |
+
"eval_groundtruth_f1_factual": 0.7157702918851219,
|
| 1123 |
+
"eval_icace_cosine": 0.47242775559425354,
|
| 1124 |
+
"eval_icace_l2": 0.6650809049606323,
|
| 1125 |
+
"eval_icace_normdiff": 0.4171872138977051,
|
| 1126 |
+
"eval_loss": 8.604122161865234,
|
| 1127 |
+
"eval_runtime": 12.9963,
|
| 1128 |
+
"eval_samples_per_second": 299.932,
|
| 1129 |
+
"eval_steps_per_second": 2.385,
|
| 1130 |
+
"step": 1700
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.95,
|
| 1134 |
+
"learning_rate": 3.136915077989602e-05,
|
| 1135 |
+
"loss": 8.1394,
|
| 1136 |
+
"step": 1720
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.98,
|
| 1140 |
+
"learning_rate": 3.1152512998266895e-05,
|
| 1141 |
+
"loss": 8.1326,
|
| 1142 |
+
"step": 1740
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 3.0,
|
| 1146 |
+
"eval_distillation_accuracy_counterfactual": 0.5328373524884557,
|
| 1147 |
+
"eval_distillation_accuracy_factual": 0.8496664956387892,
|
| 1148 |
+
"eval_distillation_f1_counterfactual": 0.466763064256023,
|
| 1149 |
+
"eval_distillation_f1_factual": 0.8351501069683767,
|
| 1150 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5230887634684454,
|
| 1151 |
+
"eval_groundtruth_f1_counterfactual": 0.46309383057125386,
|
| 1152 |
+
"eval_groundtruth_f1_factual": 0.7189084641368721,
|
| 1153 |
+
"eval_icace_cosine": 0.47281646728515625,
|
| 1154 |
+
"eval_icace_l2": 0.6447048783302307,
|
| 1155 |
+
"eval_icace_normdiff": 0.3949274718761444,
|
| 1156 |
+
"eval_loss": 8.471474647521973,
|
| 1157 |
+
"eval_runtime": 12.9383,
|
| 1158 |
+
"eval_samples_per_second": 301.276,
|
| 1159 |
+
"eval_steps_per_second": 2.396,
|
| 1160 |
+
"step": 1750
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 3.02,
|
| 1164 |
+
"learning_rate": 3.0935875216637785e-05,
|
| 1165 |
+
"loss": 8.0867,
|
| 1166 |
+
"step": 1760
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 3.05,
|
| 1170 |
+
"learning_rate": 3.071923743500867e-05,
|
| 1171 |
+
"loss": 8.2012,
|
| 1172 |
+
"step": 1780
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 3.09,
|
| 1176 |
+
"learning_rate": 3.0502599653379547e-05,
|
| 1177 |
+
"loss": 8.1176,
|
| 1178 |
+
"step": 1800
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.09,
|
| 1182 |
+
"eval_distillation_accuracy_counterfactual": 0.5320677270395074,
|
| 1183 |
+
"eval_distillation_accuracy_factual": 0.8473576192919446,
|
| 1184 |
+
"eval_distillation_f1_counterfactual": 0.44372130842528623,
|
| 1185 |
+
"eval_distillation_f1_factual": 0.8338491634681494,
|
| 1186 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5076962544894817,
|
| 1187 |
+
"eval_groundtruth_f1_counterfactual": 0.4271300744187679,
|
| 1188 |
+
"eval_groundtruth_f1_factual": 0.7089616826609815,
|
| 1189 |
+
"eval_icace_cosine": 0.4722249507904053,
|
| 1190 |
+
"eval_icace_l2": 0.65628582239151,
|
| 1191 |
+
"eval_icace_normdiff": 0.40942007303237915,
|
| 1192 |
+
"eval_loss": 8.602928161621094,
|
| 1193 |
+
"eval_runtime": 12.9577,
|
| 1194 |
+
"eval_samples_per_second": 300.825,
|
| 1195 |
+
"eval_steps_per_second": 2.392,
|
| 1196 |
+
"step": 1800
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 3.12,
|
| 1200 |
+
"learning_rate": 3.0285961871750434e-05,
|
| 1201 |
+
"loss": 8.1985,
|
| 1202 |
+
"step": 1820
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 3.16,
|
| 1206 |
+
"learning_rate": 3.006932409012132e-05,
|
| 1207 |
+
"loss": 8.2358,
|
| 1208 |
+
"step": 1840
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 3.17,
|
| 1212 |
+
"eval_distillation_accuracy_counterfactual": 0.5179579271421242,
|
| 1213 |
+
"eval_distillation_accuracy_factual": 0.8463314520266804,
|
| 1214 |
+
"eval_distillation_f1_counterfactual": 0.4547140810789615,
|
| 1215 |
+
"eval_distillation_f1_factual": 0.8318522122913702,
|
| 1216 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5179579271421242,
|
| 1217 |
+
"eval_groundtruth_f1_counterfactual": 0.46111342311742015,
|
| 1218 |
+
"eval_groundtruth_f1_factual": 0.702200862113018,
|
| 1219 |
+
"eval_icace_cosine": 0.47784164547920227,
|
| 1220 |
+
"eval_icace_l2": 0.6577954292297363,
|
| 1221 |
+
"eval_icace_normdiff": 0.40519601106643677,
|
| 1222 |
+
"eval_loss": 8.596319198608398,
|
| 1223 |
+
"eval_runtime": 13.0549,
|
| 1224 |
+
"eval_samples_per_second": 298.584,
|
| 1225 |
+
"eval_steps_per_second": 2.375,
|
| 1226 |
+
"step": 1850
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.19,
|
| 1230 |
+
"learning_rate": 2.98526863084922e-05,
|
| 1231 |
+
"loss": 8.1927,
|
| 1232 |
+
"step": 1860
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 3.22,
|
| 1236 |
+
"learning_rate": 2.9636048526863086e-05,
|
| 1237 |
+
"loss": 8.0759,
|
| 1238 |
+
"step": 1880
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 3.26,
|
| 1242 |
+
"learning_rate": 2.9419410745233972e-05,
|
| 1243 |
+
"loss": 8.0023,
|
| 1244 |
+
"step": 1900
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 3.26,
|
| 1248 |
+
"eval_distillation_accuracy_counterfactual": 0.5410466906105695,
|
| 1249 |
+
"eval_distillation_accuracy_factual": 0.8442791174961519,
|
| 1250 |
+
"eval_distillation_f1_counterfactual": 0.4708691905840775,
|
| 1251 |
+
"eval_distillation_f1_factual": 0.8314060332226945,
|
| 1252 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5287326834273987,
|
| 1253 |
+
"eval_groundtruth_f1_counterfactual": 0.4662134522268956,
|
| 1254 |
+
"eval_groundtruth_f1_factual": 0.7036825171946501,
|
| 1255 |
+
"eval_icace_cosine": 0.46827077865600586,
|
| 1256 |
+
"eval_icace_l2": 0.6418067812919617,
|
| 1257 |
+
"eval_icace_normdiff": 0.39047762751579285,
|
| 1258 |
+
"eval_loss": 8.510624885559082,
|
| 1259 |
+
"eval_runtime": 12.9429,
|
| 1260 |
+
"eval_samples_per_second": 301.169,
|
| 1261 |
+
"eval_steps_per_second": 2.395,
|
| 1262 |
+
"step": 1900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.29,
|
| 1266 |
+
"learning_rate": 2.920277296360485e-05,
|
| 1267 |
+
"loss": 8.0164,
|
| 1268 |
+
"step": 1920
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.33,
|
| 1272 |
+
"learning_rate": 2.8986135181975738e-05,
|
| 1273 |
+
"loss": 8.0343,
|
| 1274 |
+
"step": 1940
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 3.34,
|
| 1278 |
+
"eval_distillation_accuracy_counterfactual": 0.5356593124679323,
|
| 1279 |
+
"eval_distillation_accuracy_factual": 0.8594150846587993,
|
| 1280 |
+
"eval_distillation_f1_counterfactual": 0.4646747781692526,
|
| 1281 |
+
"eval_distillation_f1_factual": 0.8458748080305407,
|
| 1282 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5230887634684454,
|
| 1283 |
+
"eval_groundtruth_f1_counterfactual": 0.4587518415854972,
|
| 1284 |
+
"eval_groundtruth_f1_factual": 0.725017307995671,
|
| 1285 |
+
"eval_icace_cosine": 0.46998363733291626,
|
| 1286 |
+
"eval_icace_l2": 0.6633685231208801,
|
| 1287 |
+
"eval_icace_normdiff": 0.410533607006073,
|
| 1288 |
+
"eval_loss": 8.549391746520996,
|
| 1289 |
+
"eval_runtime": 13.0331,
|
| 1290 |
+
"eval_samples_per_second": 299.085,
|
| 1291 |
+
"eval_steps_per_second": 2.379,
|
| 1292 |
+
"step": 1950
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 3.36,
|
| 1296 |
+
"learning_rate": 2.8769497400346624e-05,
|
| 1297 |
+
"loss": 8.1109,
|
| 1298 |
+
"step": 1960
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 3.4,
|
| 1302 |
+
"learning_rate": 2.8552859618717503e-05,
|
| 1303 |
+
"loss": 7.8661,
|
| 1304 |
+
"step": 1980
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.43,
|
| 1308 |
+
"learning_rate": 2.833622183708839e-05,
|
| 1309 |
+
"loss": 7.9558,
|
| 1310 |
+
"step": 2000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.43,
|
| 1314 |
+
"eval_distillation_accuracy_counterfactual": 0.5338635197537198,
|
| 1315 |
+
"eval_distillation_accuracy_factual": 0.8483837865572088,
|
| 1316 |
+
"eval_distillation_f1_counterfactual": 0.46569718446185726,
|
| 1317 |
+
"eval_distillation_f1_factual": 0.8346114011854626,
|
| 1318 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5269368907131863,
|
| 1319 |
+
"eval_groundtruth_f1_counterfactual": 0.46415122006349785,
|
| 1320 |
+
"eval_groundtruth_f1_factual": 0.7156362232258703,
|
| 1321 |
+
"eval_icace_cosine": 0.46725159883499146,
|
| 1322 |
+
"eval_icace_l2": 0.660638153553009,
|
| 1323 |
+
"eval_icace_normdiff": 0.40762051939964294,
|
| 1324 |
+
"eval_loss": 8.534335136413574,
|
| 1325 |
+
"eval_runtime": 12.9678,
|
| 1326 |
+
"eval_samples_per_second": 300.591,
|
| 1327 |
+
"eval_steps_per_second": 2.391,
|
| 1328 |
+
"step": 2000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.46,
|
| 1332 |
+
"learning_rate": 2.8119584055459276e-05,
|
| 1333 |
+
"loss": 8.0163,
|
| 1334 |
+
"step": 2020
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 3.5,
|
| 1338 |
+
"learning_rate": 2.7902946273830156e-05,
|
| 1339 |
+
"loss": 8.1543,
|
| 1340 |
+
"step": 2040
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.52,
|
| 1344 |
+
"eval_distillation_accuracy_counterfactual": 0.5184710107747563,
|
| 1345 |
+
"eval_distillation_accuracy_factual": 0.8417136993329912,
|
| 1346 |
+
"eval_distillation_f1_counterfactual": 0.45048032767829593,
|
| 1347 |
+
"eval_distillation_f1_factual": 0.8277337474547333,
|
| 1348 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5187275525910724,
|
| 1349 |
+
"eval_groundtruth_f1_counterfactual": 0.4579942611753098,
|
| 1350 |
+
"eval_groundtruth_f1_factual": 0.7262909345250802,
|
| 1351 |
+
"eval_icace_cosine": 0.47337350249290466,
|
| 1352 |
+
"eval_icace_l2": 0.6622626781463623,
|
| 1353 |
+
"eval_icace_normdiff": 0.40650057792663574,
|
| 1354 |
+
"eval_loss": 8.670907020568848,
|
| 1355 |
+
"eval_runtime": 12.913,
|
| 1356 |
+
"eval_samples_per_second": 301.867,
|
| 1357 |
+
"eval_steps_per_second": 2.401,
|
| 1358 |
+
"step": 2050
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 3.53,
|
| 1362 |
+
"learning_rate": 2.7686308492201042e-05,
|
| 1363 |
+
"loss": 8.0555,
|
| 1364 |
+
"step": 2060
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 3.57,
|
| 1368 |
+
"learning_rate": 2.7469670710571928e-05,
|
| 1369 |
+
"loss": 8.003,
|
| 1370 |
+
"step": 2080
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.6,
|
| 1374 |
+
"learning_rate": 2.7253032928942808e-05,
|
| 1375 |
+
"loss": 8.1222,
|
| 1376 |
+
"step": 2100
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 3.6,
|
| 1380 |
+
"eval_distillation_accuracy_counterfactual": 0.538481272447409,
|
| 1381 |
+
"eval_distillation_accuracy_factual": 0.8324781939456132,
|
| 1382 |
+
"eval_distillation_f1_counterfactual": 0.45321151762235823,
|
| 1383 |
+
"eval_distillation_f1_factual": 0.816404323283801,
|
| 1384 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5256541816316059,
|
| 1385 |
+
"eval_groundtruth_f1_counterfactual": 0.4475943574688602,
|
| 1386 |
+
"eval_groundtruth_f1_factual": 0.7218836713261483,
|
| 1387 |
+
"eval_icace_cosine": 0.47021886706352234,
|
| 1388 |
+
"eval_icace_l2": 0.6517447233200073,
|
| 1389 |
+
"eval_icace_normdiff": 0.403333842754364,
|
| 1390 |
+
"eval_loss": 8.577171325683594,
|
| 1391 |
+
"eval_runtime": 12.9514,
|
| 1392 |
+
"eval_samples_per_second": 300.97,
|
| 1393 |
+
"eval_steps_per_second": 2.394,
|
| 1394 |
+
"step": 2100
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.64,
|
| 1398 |
+
"learning_rate": 2.7036395147313694e-05,
|
| 1399 |
+
"loss": 8.2765,
|
| 1400 |
+
"step": 2120
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 3.67,
|
| 1404 |
+
"learning_rate": 2.6819757365684577e-05,
|
| 1405 |
+
"loss": 8.0517,
|
| 1406 |
+
"step": 2140
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 3.69,
|
| 1410 |
+
"eval_distillation_accuracy_counterfactual": 0.5318111852231914,
|
| 1411 |
+
"eval_distillation_accuracy_factual": 0.8437660338635198,
|
| 1412 |
+
"eval_distillation_f1_counterfactual": 0.4762348812215591,
|
| 1413 |
+
"eval_distillation_f1_factual": 0.8278687320161431,
|
| 1414 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5200102616726526,
|
| 1415 |
+
"eval_groundtruth_f1_counterfactual": 0.47065260912789525,
|
| 1416 |
+
"eval_groundtruth_f1_factual": 0.7123545192728437,
|
| 1417 |
+
"eval_icace_cosine": 0.47268736362457275,
|
| 1418 |
+
"eval_icace_l2": 0.6466090083122253,
|
| 1419 |
+
"eval_icace_normdiff": 0.4019717872142792,
|
| 1420 |
+
"eval_loss": 8.487833023071289,
|
| 1421 |
+
"eval_runtime": 12.9963,
|
| 1422 |
+
"eval_samples_per_second": 299.933,
|
| 1423 |
+
"eval_steps_per_second": 2.385,
|
| 1424 |
+
"step": 2150
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 3.7,
|
| 1428 |
+
"learning_rate": 2.660311958405546e-05,
|
| 1429 |
+
"loss": 8.1498,
|
| 1430 |
+
"step": 2160
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.74,
|
| 1434 |
+
"learning_rate": 2.6386481802426343e-05,
|
| 1435 |
+
"loss": 8.1688,
|
| 1436 |
+
"step": 2180
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.77,
|
| 1440 |
+
"learning_rate": 2.616984402079723e-05,
|
| 1441 |
+
"loss": 7.9705,
|
| 1442 |
+
"step": 2200
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 3.77,
|
| 1446 |
+
"eval_distillation_accuracy_counterfactual": 0.5377116469984607,
|
| 1447 |
+
"eval_distillation_accuracy_factual": 0.8240123140071832,
|
| 1448 |
+
"eval_distillation_f1_counterfactual": 0.4741425223222236,
|
| 1449 |
+
"eval_distillation_f1_factual": 0.8055011349134608,
|
| 1450 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5269368907131863,
|
| 1451 |
+
"eval_groundtruth_f1_counterfactual": 0.4693711263304287,
|
| 1452 |
+
"eval_groundtruth_f1_factual": 0.7106558328588921,
|
| 1453 |
+
"eval_icace_cosine": 0.4721924662590027,
|
| 1454 |
+
"eval_icace_l2": 0.6477442383766174,
|
| 1455 |
+
"eval_icace_normdiff": 0.39685875177383423,
|
| 1456 |
+
"eval_loss": 8.761940956115723,
|
| 1457 |
+
"eval_runtime": 12.9582,
|
| 1458 |
+
"eval_samples_per_second": 300.813,
|
| 1459 |
+
"eval_steps_per_second": 2.392,
|
| 1460 |
+
"step": 2200
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 3.81,
|
| 1464 |
+
"learning_rate": 2.595320623916811e-05,
|
| 1465 |
+
"loss": 7.943,
|
| 1466 |
+
"step": 2220
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 3.84,
|
| 1470 |
+
"learning_rate": 2.5736568457538995e-05,
|
| 1471 |
+
"loss": 8.1214,
|
| 1472 |
+
"step": 2240
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.86,
|
| 1476 |
+
"eval_distillation_accuracy_counterfactual": 0.5348896870189841,
|
| 1477 |
+
"eval_distillation_accuracy_factual": 0.8435094920472037,
|
| 1478 |
+
"eval_distillation_f1_counterfactual": 0.46585115605288197,
|
| 1479 |
+
"eval_distillation_f1_factual": 0.8294546068203583,
|
| 1480 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5207798871216008,
|
| 1481 |
+
"eval_groundtruth_f1_counterfactual": 0.45831312719861905,
|
| 1482 |
+
"eval_groundtruth_f1_factual": 0.7120266943001898,
|
| 1483 |
+
"eval_icace_cosine": 0.4707246422767639,
|
| 1484 |
+
"eval_icace_l2": 0.6443170309066772,
|
| 1485 |
+
"eval_icace_normdiff": 0.3977286219596863,
|
| 1486 |
+
"eval_loss": 8.511881828308105,
|
| 1487 |
+
"eval_runtime": 13.0295,
|
| 1488 |
+
"eval_samples_per_second": 299.167,
|
| 1489 |
+
"eval_steps_per_second": 2.379,
|
| 1490 |
+
"step": 2250
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 3.88,
|
| 1494 |
+
"learning_rate": 2.551993067590988e-05,
|
| 1495 |
+
"loss": 7.963,
|
| 1496 |
+
"step": 2260
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.91,
|
| 1500 |
+
"learning_rate": 2.530329289428076e-05,
|
| 1501 |
+
"loss": 7.9637,
|
| 1502 |
+
"step": 2280
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 3.95,
|
| 1506 |
+
"learning_rate": 2.5086655112651647e-05,
|
| 1507 |
+
"loss": 8.1591,
|
| 1508 |
+
"step": 2300
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 3.95,
|
| 1512 |
+
"eval_distillation_accuracy_counterfactual": 0.5092355053873782,
|
| 1513 |
+
"eval_distillation_accuracy_factual": 0.8486403283735249,
|
| 1514 |
+
"eval_distillation_f1_counterfactual": 0.42695574999404207,
|
| 1515 |
+
"eval_distillation_f1_factual": 0.8334085301941027,
|
| 1516 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5025654181631606,
|
| 1517 |
+
"eval_groundtruth_f1_counterfactual": 0.4290408562754823,
|
| 1518 |
+
"eval_groundtruth_f1_factual": 0.7225390626347794,
|
| 1519 |
+
"eval_icace_cosine": 0.4755522310733795,
|
| 1520 |
+
"eval_icace_l2": 0.6715899109840393,
|
| 1521 |
+
"eval_icace_normdiff": 0.42764294147491455,
|
| 1522 |
+
"eval_loss": 8.715620994567871,
|
| 1523 |
+
"eval_runtime": 12.9596,
|
| 1524 |
+
"eval_samples_per_second": 300.781,
|
| 1525 |
+
"eval_steps_per_second": 2.392,
|
| 1526 |
+
"step": 2300
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 3.98,
|
| 1530 |
+
"learning_rate": 2.4870017331022533e-05,
|
| 1531 |
+
"loss": 8.0792,
|
| 1532 |
+
"step": 2320
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 4.01,
|
| 1536 |
+
"learning_rate": 2.4653379549393416e-05,
|
| 1537 |
+
"loss": 7.8229,
|
| 1538 |
+
"step": 2340
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 4.03,
|
| 1542 |
+
"eval_distillation_accuracy_counterfactual": 0.5531041559774243,
|
| 1543 |
+
"eval_distillation_accuracy_factual": 0.8340174448435095,
|
| 1544 |
+
"eval_distillation_f1_counterfactual": 0.4729251628273044,
|
| 1545 |
+
"eval_distillation_f1_factual": 0.8182553778403516,
|
| 1546 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5364289379168804,
|
| 1547 |
+
"eval_groundtruth_f1_counterfactual": 0.4660967256300985,
|
| 1548 |
+
"eval_groundtruth_f1_factual": 0.7131629444896277,
|
| 1549 |
+
"eval_icace_cosine": 0.47335678339004517,
|
| 1550 |
+
"eval_icace_l2": 0.6358506679534912,
|
| 1551 |
+
"eval_icace_normdiff": 0.38684460520744324,
|
| 1552 |
+
"eval_loss": 8.613683700561523,
|
| 1553 |
+
"eval_runtime": 12.9404,
|
| 1554 |
+
"eval_samples_per_second": 301.226,
|
| 1555 |
+
"eval_steps_per_second": 2.396,
|
| 1556 |
+
"step": 2350
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.05,
|
| 1560 |
+
"learning_rate": 2.44367417677643e-05,
|
| 1561 |
+
"loss": 8.022,
|
| 1562 |
+
"step": 2360
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 4.08,
|
| 1566 |
+
"learning_rate": 2.4220103986135185e-05,
|
| 1567 |
+
"loss": 7.8912,
|
| 1568 |
+
"step": 2380
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 4.12,
|
| 1572 |
+
"learning_rate": 2.4003466204506068e-05,
|
| 1573 |
+
"loss": 8.0281,
|
| 1574 |
+
"step": 2400
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 4.12,
|
| 1578 |
+
"eval_distillation_accuracy_counterfactual": 0.5456644433042586,
|
| 1579 |
+
"eval_distillation_accuracy_factual": 0.848896870189841,
|
| 1580 |
+
"eval_distillation_f1_counterfactual": 0.4776857954065169,
|
| 1581 |
+
"eval_distillation_f1_factual": 0.8326616872239038,
|
| 1582 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5297588506926629,
|
| 1583 |
+
"eval_groundtruth_f1_counterfactual": 0.46905822427531607,
|
| 1584 |
+
"eval_groundtruth_f1_factual": 0.7201395082994183,
|
| 1585 |
+
"eval_icace_cosine": 0.4664005935192108,
|
| 1586 |
+
"eval_icace_l2": 0.6372502446174622,
|
| 1587 |
+
"eval_icace_normdiff": 0.39846205711364746,
|
| 1588 |
+
"eval_loss": 8.315762519836426,
|
| 1589 |
+
"eval_runtime": 13.0359,
|
| 1590 |
+
"eval_samples_per_second": 299.021,
|
| 1591 |
+
"eval_steps_per_second": 2.378,
|
| 1592 |
+
"step": 2400
|
| 1593 |
+
}
|
| 1594 |
+
],
|
| 1595 |
+
"max_steps": 4616,
|
| 1596 |
+
"num_train_epochs": 8,
|
| 1597 |
+
"total_flos": 2.0194317434112e+16,
|
| 1598 |
+
"trial_name": null,
|
| 1599 |
+
"trial_params": null
|
| 1600 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01cd354140850926a3f867841170bfd37cc80b1dd306bf33b218250be6503d7f
|
| 3 |
+
size 3183
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|