Hugging Face's logo

pharaouk
/
loki5

Model card Files
xet
 Community
Farouk commited on
Commit
1524afd
·
1 Parent(s): b662a94

Training in progress, step 600

Browse files
Files changed (16) hide show
  1. adapter_model.bin +1 -1
  2. checkpoint-400/adapter_model/adapter_model/README.md +20 -0
  3. checkpoint-400/adapter_model/adapter_model/adapter_config.json +26 -0
  4. checkpoint-400/adapter_model/adapter_model/adapter_model.bin +3 -0
  5. checkpoint-600/README.md +20 -0
  6. checkpoint-600/adapter_config.json +26 -0
  7. checkpoint-600/adapter_model.bin +3 -0
  8. checkpoint-600/added_tokens.json +3 -0
  9. checkpoint-600/optimizer.pt +3 -0
  10. checkpoint-600/rng_state.pth +3 -0
  11. checkpoint-600/scheduler.pt +3 -0
  12. checkpoint-600/special_tokens_map.json +6 -0
  13. checkpoint-600/tokenizer.model +3 -0
  14. checkpoint-600/tokenizer_config.json +35 -0
  15. checkpoint-600/trainer_state.json +4168 -0
  16. checkpoint-600/training_args.bin +3 -0
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:15dc3d4022189fb7f87c80c8a1361d8e582fea1ed0fa87bd56bcacd2aefee2ac
3
  size 319977229
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c89e58bd2273d1f3aec96cadb5ad1335fd3030873ecd8bdac3e6c2349e634cdc
3
  size 319977229
checkpoint-400/adapter_model/adapter_model/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-400/adapter_model/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16.0,
11
+ "lora_dropout": 0.4,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "k_proj",
19
+ "gate_proj",
20
+ "down_proj",
21
+ "q_proj",
22
+ "up_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-400/adapter_model/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15dc3d4022189fb7f87c80c8a1361d8e582fea1ed0fa87bd56bcacd2aefee2ac
3
+ size 319977229
checkpoint-600/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-600/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16.0,
11
+ "lora_dropout": 0.4,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "k_proj",
19
+ "gate_proj",
20
+ "down_proj",
21
+ "q_proj",
22
+ "up_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-600/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c89e58bd2273d1f3aec96cadb5ad1335fd3030873ecd8bdac3e6c2349e634cdc
3
+ size 319977229
checkpoint-600/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<pad>": 32000
3
+ }
checkpoint-600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d156b0d51e4b0baa73f3d06fde4466d47d0c315e77916be4e9f4b97e5603e24c
3
+ size 1279539973
checkpoint-600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:428800511b314d28f9b52a9e7f3e52507c9b3807682a1f0b719700359060bcc4
3
+ size 14511
checkpoint-600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:224d0a6b8f8e758a7d5e51c51008af46238ffc13846e68ad246d75ce12bc59ff
3
+ size 627
checkpoint-600/special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "<unk>",
5
+ "unk_token": "<unk>"
6
+ }
checkpoint-600/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
checkpoint-600/tokenizer_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": false,
13
+ "eos_token": {
14
+ "__type": "AddedToken",
15
+ "content": "</s>",
16
+ "lstrip": false,
17
+ "normalized": true,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "legacy": false,
22
+ "model_max_length": 1000000000000000019884624838656,
23
+ "pad_token": null,
24
+ "padding_side": "right",
25
+ "sp_model_kwargs": {},
26
+ "tokenizer_class": "LlamaTokenizer",
27
+ "unk_token": {
28
+ "__type": "AddedToken",
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ }
35
+ }
checkpoint-600/trainer_state.json ADDED
@@ -0,0 +1,4168 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.510773241519928,
3
+ "best_model_checkpoint": "./output_v2/7b_cluster00_Nous-Hermes-llama-2-7b_codellama_blob_1/checkpoint-600",
4
+ "epoch": 1.6842105263157894,
5
+ "global_step": 600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 2e-06,
13
+ "loss": 0.7795,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4e-06,
19
+ "loss": 0.6789,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 5.999999999999999e-06,
25
+ "loss": 0.6847,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 8e-06,
31
+ "loss": 0.6876,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 9.999999999999999e-06,
37
+ "loss": 0.5964,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 1.1999999999999999e-05,
43
+ "loss": 0.6592,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 1.4e-05,
49
+ "loss": 0.6261,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 1.6e-05,
55
+ "loss": 0.6892,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 1.7999999999999997e-05,
61
+ "loss": 0.538,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 0.03,
66
+ "learning_rate": 1.9999999999999998e-05,
67
+ "loss": 0.6015,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 2.2e-05,
73
+ "loss": 0.7061,
74
+ "step": 11
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 2.3999999999999997e-05,
79
+ "loss": 0.6329,
80
+ "step": 12
81
+ },
82
+ {
83
+ "epoch": 0.04,
84
+ "learning_rate": 2.6e-05,
85
+ "loss": 0.5633,
86
+ "step": 13
87
+ },
88
+ {
89
+ "epoch": 0.04,
90
+ "learning_rate": 2.8e-05,
91
+ "loss": 0.5499,
92
+ "step": 14
93
+ },
94
+ {
95
+ "epoch": 0.04,
96
+ "learning_rate": 2.9999999999999997e-05,
97
+ "loss": 0.5846,
98
+ "step": 15
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 3.2e-05,
103
+ "loss": 0.7056,
104
+ "step": 16
105
+ },
106
+ {
107
+ "epoch": 0.05,
108
+ "learning_rate": 3.399999999999999e-05,
109
+ "loss": 0.9478,
110
+ "step": 17
111
+ },
112
+ {
113
+ "epoch": 0.05,
114
+ "learning_rate": 3.5999999999999994e-05,
115
+ "loss": 0.7119,
116
+ "step": 18
117
+ },
118
+ {
119
+ "epoch": 0.05,
120
+ "learning_rate": 3.8e-05,
121
+ "loss": 0.5115,
122
+ "step": 19
123
+ },
124
+ {
125
+ "epoch": 0.06,
126
+ "learning_rate": 3.9999999999999996e-05,
127
+ "loss": 0.6481,
128
+ "step": 20
129
+ },
130
+ {
131
+ "epoch": 0.06,
132
+ "learning_rate": 4.2e-05,
133
+ "loss": 1.0479,
134
+ "step": 21
135
+ },
136
+ {
137
+ "epoch": 0.06,
138
+ "learning_rate": 4.4e-05,
139
+ "loss": 0.9266,
140
+ "step": 22
141
+ },
142
+ {
143
+ "epoch": 0.06,
144
+ "learning_rate": 4.599999999999999e-05,
145
+ "loss": 0.7517,
146
+ "step": 23
147
+ },
148
+ {
149
+ "epoch": 0.07,
150
+ "learning_rate": 4.7999999999999994e-05,
151
+ "loss": 0.7024,
152
+ "step": 24
153
+ },
154
+ {
155
+ "epoch": 0.07,
156
+ "learning_rate": 4.9999999999999996e-05,
157
+ "loss": 0.6996,
158
+ "step": 25
159
+ },
160
+ {
161
+ "epoch": 0.07,
162
+ "eval_loss": 0.6741535663604736,
163
+ "eval_runtime": 26.372,
164
+ "eval_samples_per_second": 7.584,
165
+ "eval_steps_per_second": 1.896,
166
+ "step": 25
167
+ },
168
+ {
169
+ "dharma_eval_accuracy": 0.47090526975371527,
170
+ "dharma_eval_accuracy_ARC-Challenge": 0.6296296296296297,
171
+ "dharma_eval_accuracy_ARC-Easy": 0.7592592592592593,
172
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
173
+ "dharma_eval_accuracy_MMLU": 0.4444444444444444,
174
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
175
+ "dharma_eval_accuracy_bigbench": 0.4098360655737705,
176
+ "dharma_eval_accuracy_openbookqa": 0.2962962962962963,
177
+ "dharma_eval_accuracy_truthful_qa": 0.37037037037037035,
178
+ "dharma_eval_accuracy_winogrande": 0.37037037037037035,
179
+ "dharma_loss": 3.7834435443878176,
180
+ "epoch": 0.07,
181
+ "step": 25
182
+ },
183
+ {
184
+ "epoch": 0.07,
185
+ "learning_rate": 5.2e-05,
186
+ "loss": 0.6123,
187
+ "step": 26
188
+ },
189
+ {
190
+ "epoch": 0.08,
191
+ "learning_rate": 5.399999999999999e-05,
192
+ "loss": 0.608,
193
+ "step": 27
194
+ },
195
+ {
196
+ "epoch": 0.08,
197
+ "learning_rate": 5.6e-05,
198
+ "loss": 0.4845,
199
+ "step": 28
200
+ },
201
+ {
202
+ "epoch": 0.08,
203
+ "learning_rate": 5.7999999999999994e-05,
204
+ "loss": 0.6026,
205
+ "step": 29
206
+ },
207
+ {
208
+ "epoch": 0.08,
209
+ "learning_rate": 5.9999999999999995e-05,
210
+ "loss": 0.87,
211
+ "step": 30
212
+ },
213
+ {
214
+ "epoch": 0.09,
215
+ "learning_rate": 6.199999999999999e-05,
216
+ "loss": 0.9898,
217
+ "step": 31
218
+ },
219
+ {
220
+ "epoch": 0.09,
221
+ "learning_rate": 6.4e-05,
222
+ "loss": 0.7354,
223
+ "step": 32
224
+ },
225
+ {
226
+ "epoch": 0.09,
227
+ "learning_rate": 6.599999999999999e-05,
228
+ "loss": 0.5053,
229
+ "step": 33
230
+ },
231
+ {
232
+ "epoch": 0.1,
233
+ "learning_rate": 6.799999999999999e-05,
234
+ "loss": 0.6644,
235
+ "step": 34
236
+ },
237
+ {
238
+ "epoch": 0.1,
239
+ "learning_rate": 7e-05,
240
+ "loss": 0.6932,
241
+ "step": 35
242
+ },
243
+ {
244
+ "epoch": 0.1,
245
+ "learning_rate": 7.199999999999999e-05,
246
+ "loss": 0.7425,
247
+ "step": 36
248
+ },
249
+ {
250
+ "epoch": 0.1,
251
+ "learning_rate": 7.4e-05,
252
+ "loss": 0.8268,
253
+ "step": 37
254
+ },
255
+ {
256
+ "epoch": 0.11,
257
+ "learning_rate": 7.6e-05,
258
+ "loss": 0.5957,
259
+ "step": 38
260
+ },
261
+ {
262
+ "epoch": 0.11,
263
+ "learning_rate": 7.8e-05,
264
+ "loss": 0.5579,
265
+ "step": 39
266
+ },
267
+ {
268
+ "epoch": 0.11,
269
+ "learning_rate": 7.999999999999999e-05,
270
+ "loss": 0.7978,
271
+ "step": 40
272
+ },
273
+ {
274
+ "epoch": 0.12,
275
+ "learning_rate": 8.199999999999999e-05,
276
+ "loss": 0.8261,
277
+ "step": 41
278
+ },
279
+ {
280
+ "epoch": 0.12,
281
+ "learning_rate": 8.4e-05,
282
+ "loss": 0.7618,
283
+ "step": 42
284
+ },
285
+ {
286
+ "epoch": 0.12,
287
+ "learning_rate": 8.6e-05,
288
+ "loss": 0.6062,
289
+ "step": 43
290
+ },
291
+ {
292
+ "epoch": 0.12,
293
+ "learning_rate": 8.8e-05,
294
+ "loss": 0.5791,
295
+ "step": 44
296
+ },
297
+ {
298
+ "epoch": 0.13,
299
+ "learning_rate": 8.999999999999999e-05,
300
+ "loss": 0.6008,
301
+ "step": 45
302
+ },
303
+ {
304
+ "epoch": 0.13,
305
+ "learning_rate": 9.199999999999999e-05,
306
+ "loss": 0.5143,
307
+ "step": 46
308
+ },
309
+ {
310
+ "epoch": 0.13,
311
+ "learning_rate": 9.4e-05,
312
+ "loss": 0.5329,
313
+ "step": 47
314
+ },
315
+ {
316
+ "epoch": 0.13,
317
+ "learning_rate": 9.599999999999999e-05,
318
+ "loss": 0.6973,
319
+ "step": 48
320
+ },
321
+ {
322
+ "epoch": 0.14,
323
+ "learning_rate": 9.799999999999998e-05,
324
+ "loss": 0.615,
325
+ "step": 49
326
+ },
327
+ {
328
+ "epoch": 0.14,
329
+ "learning_rate": 9.999999999999999e-05,
330
+ "loss": 0.6362,
331
+ "step": 50
332
+ },
333
+ {
334
+ "epoch": 0.14,
335
+ "eval_loss": 0.6042094826698303,
336
+ "eval_runtime": 26.344,
337
+ "eval_samples_per_second": 7.592,
338
+ "eval_steps_per_second": 1.898,
339
+ "step": 50
340
+ },
341
+ {
342
+ "dharma_eval_accuracy": 0.4825425901342051,
343
+ "dharma_eval_accuracy_ARC-Challenge": 0.6666666666666666,
344
+ "dharma_eval_accuracy_ARC-Easy": 0.6851851851851852,
345
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
346
+ "dharma_eval_accuracy_MMLU": 0.42592592592592593,
347
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
348
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
349
+ "dharma_eval_accuracy_openbookqa": 0.3333333333333333,
350
+ "dharma_eval_accuracy_truthful_qa": 0.35185185185185186,
351
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
352
+ "dharma_loss": 2.1548793907165527,
353
+ "epoch": 0.14,
354
+ "step": 50
355
+ },
356
+ {
357
+ "epoch": 0.14,
358
+ "learning_rate": 0.000102,
359
+ "loss": 0.6567,
360
+ "step": 51
361
+ },
362
+ {
363
+ "epoch": 0.15,
364
+ "learning_rate": 0.000104,
365
+ "loss": 0.9786,
366
+ "step": 52
367
+ },
368
+ {
369
+ "epoch": 0.15,
370
+ "learning_rate": 0.00010599999999999999,
371
+ "loss": 0.6146,
372
+ "step": 53
373
+ },
374
+ {
375
+ "epoch": 0.15,
376
+ "learning_rate": 0.00010799999999999998,
377
+ "loss": 0.6402,
378
+ "step": 54
379
+ },
380
+ {
381
+ "epoch": 0.15,
382
+ "learning_rate": 0.00010999999999999998,
383
+ "loss": 0.5906,
384
+ "step": 55
385
+ },
386
+ {
387
+ "epoch": 0.16,
388
+ "learning_rate": 0.000112,
389
+ "loss": 0.6441,
390
+ "step": 56
391
+ },
392
+ {
393
+ "epoch": 0.16,
394
+ "learning_rate": 0.00011399999999999999,
395
+ "loss": 0.6887,
396
+ "step": 57
397
+ },
398
+ {
399
+ "epoch": 0.16,
400
+ "learning_rate": 0.00011599999999999999,
401
+ "loss": 0.6283,
402
+ "step": 58
403
+ },
404
+ {
405
+ "epoch": 0.17,
406
+ "learning_rate": 0.00011799999999999998,
407
+ "loss": 0.5929,
408
+ "step": 59
409
+ },
410
+ {
411
+ "epoch": 0.17,
412
+ "learning_rate": 0.00011999999999999999,
413
+ "loss": 0.6939,
414
+ "step": 60
415
+ },
416
+ {
417
+ "epoch": 0.17,
418
+ "learning_rate": 0.000122,
419
+ "loss": 0.5535,
420
+ "step": 61
421
+ },
422
+ {
423
+ "epoch": 0.17,
424
+ "learning_rate": 0.00012399999999999998,
425
+ "loss": 0.5457,
426
+ "step": 62
427
+ },
428
+ {
429
+ "epoch": 0.18,
430
+ "learning_rate": 0.00012599999999999997,
431
+ "loss": 0.6425,
432
+ "step": 63
433
+ },
434
+ {
435
+ "epoch": 0.18,
436
+ "learning_rate": 0.000128,
437
+ "loss": 0.6855,
438
+ "step": 64
439
+ },
440
+ {
441
+ "epoch": 0.18,
442
+ "learning_rate": 0.00013,
443
+ "loss": 0.5396,
444
+ "step": 65
445
+ },
446
+ {
447
+ "epoch": 0.19,
448
+ "learning_rate": 0.00013199999999999998,
449
+ "loss": 0.6307,
450
+ "step": 66
451
+ },
452
+ {
453
+ "epoch": 0.19,
454
+ "learning_rate": 0.00013399999999999998,
455
+ "loss": 0.6396,
456
+ "step": 67
457
+ },
458
+ {
459
+ "epoch": 0.19,
460
+ "learning_rate": 0.00013599999999999997,
461
+ "loss": 0.5727,
462
+ "step": 68
463
+ },
464
+ {
465
+ "epoch": 0.19,
466
+ "learning_rate": 0.000138,
467
+ "loss": 0.4812,
468
+ "step": 69
469
+ },
470
+ {
471
+ "epoch": 0.2,
472
+ "learning_rate": 0.00014,
473
+ "loss": 0.5658,
474
+ "step": 70
475
+ },
476
+ {
477
+ "epoch": 0.2,
478
+ "learning_rate": 0.00014199999999999998,
479
+ "loss": 0.6314,
480
+ "step": 71
481
+ },
482
+ {
483
+ "epoch": 0.2,
484
+ "learning_rate": 0.00014399999999999998,
485
+ "loss": 0.6175,
486
+ "step": 72
487
+ },
488
+ {
489
+ "epoch": 0.2,
490
+ "learning_rate": 0.000146,
491
+ "loss": 0.5468,
492
+ "step": 73
493
+ },
494
+ {
495
+ "epoch": 0.21,
496
+ "learning_rate": 0.000148,
497
+ "loss": 0.6233,
498
+ "step": 74
499
+ },
500
+ {
501
+ "epoch": 0.21,
502
+ "learning_rate": 0.00015,
503
+ "loss": 0.5276,
504
+ "step": 75
505
+ },
506
+ {
507
+ "epoch": 0.21,
508
+ "eval_loss": 0.5642657279968262,
509
+ "eval_runtime": 26.2701,
510
+ "eval_samples_per_second": 7.613,
511
+ "eval_steps_per_second": 1.903,
512
+ "step": 75
513
+ },
514
+ {
515
+ "dharma_eval_accuracy": 0.45038002097061713,
516
+ "dharma_eval_accuracy_ARC-Challenge": 0.6111111111111112,
517
+ "dharma_eval_accuracy_ARC-Easy": 0.6111111111111112,
518
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
519
+ "dharma_eval_accuracy_MMLU": 0.42592592592592593,
520
+ "dharma_eval_accuracy_agieval": 0.2033898305084746,
521
+ "dharma_eval_accuracy_bigbench": 0.4426229508196721,
522
+ "dharma_eval_accuracy_openbookqa": 0.14814814814814814,
523
+ "dharma_eval_accuracy_truthful_qa": 0.42592592592592593,
524
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
525
+ "dharma_loss": 2.056876979827881,
526
+ "epoch": 0.21,
527
+ "step": 75
528
+ },
529
+ {
530
+ "epoch": 0.21,
531
+ "learning_rate": 0.000152,
532
+ "loss": 0.5308,
533
+ "step": 76
534
+ },
535
+ {
536
+ "epoch": 0.22,
537
+ "learning_rate": 0.00015399999999999998,
538
+ "loss": 0.6546,
539
+ "step": 77
540
+ },
541
+ {
542
+ "epoch": 0.22,
543
+ "learning_rate": 0.000156,
544
+ "loss": 0.6451,
545
+ "step": 78
546
+ },
547
+ {
548
+ "epoch": 0.22,
549
+ "learning_rate": 0.00015799999999999996,
550
+ "loss": 0.5255,
551
+ "step": 79
552
+ },
553
+ {
554
+ "epoch": 0.22,
555
+ "learning_rate": 0.00015999999999999999,
556
+ "loss": 0.5629,
557
+ "step": 80
558
+ },
559
+ {
560
+ "epoch": 0.23,
561
+ "learning_rate": 0.000162,
562
+ "loss": 0.6642,
563
+ "step": 81
564
+ },
565
+ {
566
+ "epoch": 0.23,
567
+ "learning_rate": 0.00016399999999999997,
568
+ "loss": 0.6393,
569
+ "step": 82
570
+ },
571
+ {
572
+ "epoch": 0.23,
573
+ "learning_rate": 0.000166,
574
+ "loss": 0.5073,
575
+ "step": 83
576
+ },
577
+ {
578
+ "epoch": 0.24,
579
+ "learning_rate": 0.000168,
580
+ "loss": 0.5628,
581
+ "step": 84
582
+ },
583
+ {
584
+ "epoch": 0.24,
585
+ "learning_rate": 0.00016999999999999999,
586
+ "loss": 0.5566,
587
+ "step": 85
588
+ },
589
+ {
590
+ "epoch": 0.24,
591
+ "learning_rate": 0.000172,
592
+ "loss": 0.4653,
593
+ "step": 86
594
+ },
595
+ {
596
+ "epoch": 0.24,
597
+ "learning_rate": 0.00017399999999999997,
598
+ "loss": 0.5322,
599
+ "step": 87
600
+ },
601
+ {
602
+ "epoch": 0.25,
603
+ "learning_rate": 0.000176,
604
+ "loss": 0.7131,
605
+ "step": 88
606
+ },
607
+ {
608
+ "epoch": 0.25,
609
+ "learning_rate": 0.000178,
610
+ "loss": 0.5586,
611
+ "step": 89
612
+ },
613
+ {
614
+ "epoch": 0.25,
615
+ "learning_rate": 0.00017999999999999998,
616
+ "loss": 0.7246,
617
+ "step": 90
618
+ },
619
+ {
620
+ "epoch": 0.26,
621
+ "learning_rate": 0.00018199999999999998,
622
+ "loss": 0.6527,
623
+ "step": 91
624
+ },
625
+ {
626
+ "epoch": 0.26,
627
+ "learning_rate": 0.00018399999999999997,
628
+ "loss": 0.8067,
629
+ "step": 92
630
+ },
631
+ {
632
+ "epoch": 0.26,
633
+ "learning_rate": 0.000186,
634
+ "loss": 0.6427,
635
+ "step": 93
636
+ },
637
+ {
638
+ "epoch": 0.26,
639
+ "learning_rate": 0.000188,
640
+ "loss": 0.6309,
641
+ "step": 94
642
+ },
643
+ {
644
+ "epoch": 0.27,
645
+ "learning_rate": 0.00018999999999999998,
646
+ "loss": 0.6111,
647
+ "step": 95
648
+ },
649
+ {
650
+ "epoch": 0.27,
651
+ "learning_rate": 0.00019199999999999998,
652
+ "loss": 0.5911,
653
+ "step": 96
654
+ },
655
+ {
656
+ "epoch": 0.27,
657
+ "learning_rate": 0.00019399999999999997,
658
+ "loss": 0.6372,
659
+ "step": 97
660
+ },
661
+ {
662
+ "epoch": 0.28,
663
+ "learning_rate": 0.00019599999999999997,
664
+ "loss": 0.5664,
665
+ "step": 98
666
+ },
667
+ {
668
+ "epoch": 0.28,
669
+ "learning_rate": 0.000198,
670
+ "loss": 0.5618,
671
+ "step": 99
672
+ },
673
+ {
674
+ "epoch": 0.28,
675
+ "learning_rate": 0.00019999999999999998,
676
+ "loss": 0.7154,
677
+ "step": 100
678
+ },
679
+ {
680
+ "epoch": 0.28,
681
+ "eval_loss": 0.5470997095108032,
682
+ "eval_runtime": 26.3144,
683
+ "eval_samples_per_second": 7.6,
684
+ "eval_steps_per_second": 1.9,
685
+ "step": 100
686
+ },
687
+ {
688
+ "dharma_eval_accuracy": 0.398478315307064,
689
+ "dharma_eval_accuracy_ARC-Challenge": 0.5740740740740741,
690
+ "dharma_eval_accuracy_ARC-Easy": 0.48148148148148145,
691
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
692
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
693
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
694
+ "dharma_eval_accuracy_bigbench": 0.4262295081967213,
695
+ "dharma_eval_accuracy_openbookqa": 0.07407407407407407,
696
+ "dharma_eval_accuracy_truthful_qa": 0.3888888888888889,
697
+ "dharma_eval_accuracy_winogrande": 0.25925925925925924,
698
+ "dharma_loss": 2.5759267053604127,
699
+ "epoch": 0.28,
700
+ "step": 100
701
+ },
702
+ {
703
+ "epoch": 0.28,
704
+ "learning_rate": 0.00020199999999999998,
705
+ "loss": 0.5837,
706
+ "step": 101
707
+ },
708
+ {
709
+ "epoch": 0.29,
710
+ "learning_rate": 0.000204,
711
+ "loss": 0.5707,
712
+ "step": 102
713
+ },
714
+ {
715
+ "epoch": 0.29,
716
+ "learning_rate": 0.00020599999999999997,
717
+ "loss": 0.7798,
718
+ "step": 103
719
+ },
720
+ {
721
+ "epoch": 0.29,
722
+ "learning_rate": 0.000208,
723
+ "loss": 0.4922,
724
+ "step": 104
725
+ },
726
+ {
727
+ "epoch": 0.29,
728
+ "learning_rate": 0.00020999999999999998,
729
+ "loss": 0.7033,
730
+ "step": 105
731
+ },
732
+ {
733
+ "epoch": 0.3,
734
+ "learning_rate": 0.00021199999999999998,
735
+ "loss": 0.7297,
736
+ "step": 106
737
+ },
738
+ {
739
+ "epoch": 0.3,
740
+ "learning_rate": 0.000214,
741
+ "loss": 0.5376,
742
+ "step": 107
743
+ },
744
+ {
745
+ "epoch": 0.3,
746
+ "learning_rate": 0.00021599999999999996,
747
+ "loss": 0.5372,
748
+ "step": 108
749
+ },
750
+ {
751
+ "epoch": 0.31,
752
+ "learning_rate": 0.00021799999999999999,
753
+ "loss": 0.7453,
754
+ "step": 109
755
+ },
756
+ {
757
+ "epoch": 0.31,
758
+ "learning_rate": 0.00021999999999999995,
759
+ "loss": 0.4582,
760
+ "step": 110
761
+ },
762
+ {
763
+ "epoch": 0.31,
764
+ "learning_rate": 0.00022199999999999998,
765
+ "loss": 0.54,
766
+ "step": 111
767
+ },
768
+ {
769
+ "epoch": 0.31,
770
+ "learning_rate": 0.000224,
771
+ "loss": 0.643,
772
+ "step": 112
773
+ },
774
+ {
775
+ "epoch": 0.32,
776
+ "learning_rate": 0.00022599999999999996,
777
+ "loss": 0.5972,
778
+ "step": 113
779
+ },
780
+ {
781
+ "epoch": 0.32,
782
+ "learning_rate": 0.00022799999999999999,
783
+ "loss": 0.5765,
784
+ "step": 114
785
+ },
786
+ {
787
+ "epoch": 0.32,
788
+ "learning_rate": 0.00023,
789
+ "loss": 0.6007,
790
+ "step": 115
791
+ },
792
+ {
793
+ "epoch": 0.33,
794
+ "learning_rate": 0.00023199999999999997,
795
+ "loss": 0.5343,
796
+ "step": 116
797
+ },
798
+ {
799
+ "epoch": 0.33,
800
+ "learning_rate": 0.000234,
801
+ "loss": 0.8235,
802
+ "step": 117
803
+ },
804
+ {
805
+ "epoch": 0.33,
806
+ "learning_rate": 0.00023599999999999996,
807
+ "loss": 0.6687,
808
+ "step": 118
809
+ },
810
+ {
811
+ "epoch": 0.33,
812
+ "learning_rate": 0.00023799999999999998,
813
+ "loss": 0.6696,
814
+ "step": 119
815
+ },
816
+ {
817
+ "epoch": 0.34,
818
+ "learning_rate": 0.00023999999999999998,
819
+ "loss": 0.77,
820
+ "step": 120
821
+ },
822
+ {
823
+ "epoch": 0.34,
824
+ "learning_rate": 0.00024199999999999997,
825
+ "loss": 0.5792,
826
+ "step": 121
827
+ },
828
+ {
829
+ "epoch": 0.34,
830
+ "learning_rate": 0.000244,
831
+ "loss": 0.3653,
832
+ "step": 122
833
+ },
834
+ {
835
+ "epoch": 0.35,
836
+ "learning_rate": 0.00024599999999999996,
837
+ "loss": 0.5344,
838
+ "step": 123
839
+ },
840
+ {
841
+ "epoch": 0.35,
842
+ "learning_rate": 0.00024799999999999996,
843
+ "loss": 0.5772,
844
+ "step": 124
845
+ },
846
+ {
847
+ "epoch": 0.35,
848
+ "learning_rate": 0.00025,
849
+ "loss": 0.7189,
850
+ "step": 125
851
+ },
852
+ {
853
+ "epoch": 0.35,
854
+ "eval_loss": 0.5514405369758606,
855
+ "eval_runtime": 26.3447,
856
+ "eval_samples_per_second": 7.592,
857
+ "eval_steps_per_second": 1.898,
858
+ "step": 125
859
+ },
860
+ {
861
+ "dharma_eval_accuracy": 0.3371043854202756,
862
+ "dharma_eval_accuracy_ARC-Challenge": 0.37037037037037035,
863
+ "dharma_eval_accuracy_ARC-Easy": 0.42592592592592593,
864
+ "dharma_eval_accuracy_BoolQ": 0.37037037037037035,
865
+ "dharma_eval_accuracy_MMLU": 0.35185185185185186,
866
+ "dharma_eval_accuracy_agieval": 0.22033898305084745,
867
+ "dharma_eval_accuracy_bigbench": 0.29508196721311475,
868
+ "dharma_eval_accuracy_openbookqa": 0.18518518518518517,
869
+ "dharma_eval_accuracy_truthful_qa": 0.3888888888888889,
870
+ "dharma_eval_accuracy_winogrande": 0.42592592592592593,
871
+ "dharma_loss": 2.1871530780792234,
872
+ "epoch": 0.35,
873
+ "step": 125
874
+ },
875
+ {
876
+ "epoch": 0.35,
877
+ "learning_rate": 0.00025199999999999995,
878
+ "loss": 0.5891,
879
+ "step": 126
880
+ },
881
+ {
882
+ "epoch": 0.36,
883
+ "learning_rate": 0.000254,
884
+ "loss": 0.8225,
885
+ "step": 127
886
+ },
887
+ {
888
+ "epoch": 0.36,
889
+ "learning_rate": 0.000256,
890
+ "loss": 0.6104,
891
+ "step": 128
892
+ },
893
+ {
894
+ "epoch": 0.36,
895
+ "learning_rate": 0.000258,
896
+ "loss": 0.6571,
897
+ "step": 129
898
+ },
899
+ {
900
+ "epoch": 0.36,
901
+ "learning_rate": 0.00026,
902
+ "loss": 0.7562,
903
+ "step": 130
904
+ },
905
+ {
906
+ "epoch": 0.37,
907
+ "learning_rate": 0.00026199999999999997,
908
+ "loss": 0.7172,
909
+ "step": 131
910
+ },
911
+ {
912
+ "epoch": 0.37,
913
+ "learning_rate": 0.00026399999999999997,
914
+ "loss": 0.4353,
915
+ "step": 132
916
+ },
917
+ {
918
+ "epoch": 0.37,
919
+ "learning_rate": 0.000266,
920
+ "loss": 0.5576,
921
+ "step": 133
922
+ },
923
+ {
924
+ "epoch": 0.38,
925
+ "learning_rate": 0.00026799999999999995,
926
+ "loss": 0.553,
927
+ "step": 134
928
+ },
929
+ {
930
+ "epoch": 0.38,
931
+ "learning_rate": 0.00027,
932
+ "loss": 0.5062,
933
+ "step": 135
934
+ },
935
+ {
936
+ "epoch": 0.38,
937
+ "learning_rate": 0.00027199999999999994,
938
+ "loss": 0.4819,
939
+ "step": 136
940
+ },
941
+ {
942
+ "epoch": 0.38,
943
+ "learning_rate": 0.000274,
944
+ "loss": 0.6195,
945
+ "step": 137
946
+ },
947
+ {
948
+ "epoch": 0.39,
949
+ "learning_rate": 0.000276,
950
+ "loss": 0.7592,
951
+ "step": 138
952
+ },
953
+ {
954
+ "epoch": 0.39,
955
+ "learning_rate": 0.000278,
956
+ "loss": 0.6118,
957
+ "step": 139
958
+ },
959
+ {
960
+ "epoch": 0.39,
961
+ "learning_rate": 0.00028,
962
+ "loss": 0.7897,
963
+ "step": 140
964
+ },
965
+ {
966
+ "epoch": 0.4,
967
+ "learning_rate": 0.00028199999999999997,
968
+ "loss": 0.5879,
969
+ "step": 141
970
+ },
971
+ {
972
+ "epoch": 0.4,
973
+ "learning_rate": 0.00028399999999999996,
974
+ "loss": 0.6055,
975
+ "step": 142
976
+ },
977
+ {
978
+ "epoch": 0.4,
979
+ "learning_rate": 0.00028599999999999996,
980
+ "loss": 0.541,
981
+ "step": 143
982
+ },
983
+ {
984
+ "epoch": 0.4,
985
+ "learning_rate": 0.00028799999999999995,
986
+ "loss": 0.7665,
987
+ "step": 144
988
+ },
989
+ {
990
+ "epoch": 0.41,
991
+ "learning_rate": 0.00029,
992
+ "loss": 0.612,
993
+ "step": 145
994
+ },
995
+ {
996
+ "epoch": 0.41,
997
+ "learning_rate": 0.000292,
998
+ "loss": 0.5397,
999
+ "step": 146
1000
+ },
1001
+ {
1002
+ "epoch": 0.41,
1003
+ "learning_rate": 0.000294,
1004
+ "loss": 0.5727,
1005
+ "step": 147
1006
+ },
1007
+ {
1008
+ "epoch": 0.42,
1009
+ "learning_rate": 0.000296,
1010
+ "loss": 1.1184,
1011
+ "step": 148
1012
+ },
1013
+ {
1014
+ "epoch": 0.42,
1015
+ "learning_rate": 0.000298,
1016
+ "loss": 0.5912,
1017
+ "step": 149
1018
+ },
1019
+ {
1020
+ "epoch": 0.42,
1021
+ "learning_rate": 0.0003,
1022
+ "loss": 0.5581,
1023
+ "step": 150
1024
+ },
1025
+ {
1026
+ "epoch": 0.42,
1027
+ "eval_loss": 0.5423073768615723,
1028
+ "eval_runtime": 26.3108,
1029
+ "eval_samples_per_second": 7.601,
1030
+ "eval_steps_per_second": 1.9,
1031
+ "step": 150
1032
+ },
1033
+ {
1034
+ "dharma_eval_accuracy": 0.3860708907481159,
1035
+ "dharma_eval_accuracy_ARC-Challenge": 0.42592592592592593,
1036
+ "dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
1037
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1038
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
1039
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
1040
+ "dharma_eval_accuracy_bigbench": 0.4426229508196721,
1041
+ "dharma_eval_accuracy_openbookqa": 0.18518518518518517,
1042
+ "dharma_eval_accuracy_truthful_qa": 0.2962962962962963,
1043
+ "dharma_eval_accuracy_winogrande": 0.35185185185185186,
1044
+ "dharma_loss": 2.471666145801544,
1045
+ "epoch": 0.42,
1046
+ "step": 150
1047
+ },
1048
+ {
1049
+ "epoch": 0.42,
1050
+ "learning_rate": 0.0002999999685313931,
1051
+ "loss": 0.656,
1052
+ "step": 151
1053
+ },
1054
+ {
1055
+ "epoch": 0.43,
1056
+ "learning_rate": 0.00029999987412558584,
1057
+ "loss": 0.5075,
1058
+ "step": 152
1059
+ },
1060
+ {
1061
+ "epoch": 0.43,
1062
+ "learning_rate": 0.0002999997167826177,
1063
+ "loss": 0.4696,
1064
+ "step": 153
1065
+ },
1066
+ {
1067
+ "epoch": 0.43,
1068
+ "learning_rate": 0.00029999949650255474,
1069
+ "loss": 0.7349,
1070
+ "step": 154
1071
+ },
1072
+ {
1073
+ "epoch": 0.44,
1074
+ "learning_rate": 0.0002999992132854894,
1075
+ "loss": 0.5935,
1076
+ "step": 155
1077
+ },
1078
+ {
1079
+ "epoch": 0.44,
1080
+ "learning_rate": 0.0002999988671315404,
1081
+ "loss": 0.4127,
1082
+ "step": 156
1083
+ },
1084
+ {
1085
+ "epoch": 0.44,
1086
+ "learning_rate": 0.0002999984580408531,
1087
+ "loss": 0.5503,
1088
+ "step": 157
1089
+ },
1090
+ {
1091
+ "epoch": 0.44,
1092
+ "learning_rate": 0.00029999798601359915,
1093
+ "loss": 0.5742,
1094
+ "step": 158
1095
+ },
1096
+ {
1097
+ "epoch": 0.45,
1098
+ "learning_rate": 0.00029999745104997654,
1099
+ "loss": 0.5606,
1100
+ "step": 159
1101
+ },
1102
+ {
1103
+ "epoch": 0.45,
1104
+ "learning_rate": 0.0002999968531502098,
1105
+ "loss": 0.6209,
1106
+ "step": 160
1107
+ },
1108
+ {
1109
+ "epoch": 0.45,
1110
+ "learning_rate": 0.0002999961923145497,
1111
+ "loss": 0.4977,
1112
+ "step": 161
1113
+ },
1114
+ {
1115
+ "epoch": 0.45,
1116
+ "learning_rate": 0.0002999954685432736,
1117
+ "loss": 0.7474,
1118
+ "step": 162
1119
+ },
1120
+ {
1121
+ "epoch": 0.46,
1122
+ "learning_rate": 0.0002999946818366852,
1123
+ "loss": 0.574,
1124
+ "step": 163
1125
+ },
1126
+ {
1127
+ "epoch": 0.46,
1128
+ "learning_rate": 0.00029999383219511444,
1129
+ "loss": 0.3991,
1130
+ "step": 164
1131
+ },
1132
+ {
1133
+ "epoch": 0.46,
1134
+ "learning_rate": 0.0002999929196189179,
1135
+ "loss": 0.6335,
1136
+ "step": 165
1137
+ },
1138
+ {
1139
+ "epoch": 0.47,
1140
+ "learning_rate": 0.0002999919441084786,
1141
+ "loss": 0.6144,
1142
+ "step": 166
1143
+ },
1144
+ {
1145
+ "epoch": 0.47,
1146
+ "learning_rate": 0.0002999909056642057,
1147
+ "loss": 0.4695,
1148
+ "step": 167
1149
+ },
1150
+ {
1151
+ "epoch": 0.47,
1152
+ "learning_rate": 0.00029998980428653496,
1153
+ "loss": 0.5322,
1154
+ "step": 168
1155
+ },
1156
+ {
1157
+ "epoch": 0.47,
1158
+ "learning_rate": 0.00029998863997592843,
1159
+ "loss": 0.6111,
1160
+ "step": 169
1161
+ },
1162
+ {
1163
+ "epoch": 0.48,
1164
+ "learning_rate": 0.00029998741273287477,
1165
+ "loss": 0.5517,
1166
+ "step": 170
1167
+ },
1168
+ {
1169
+ "epoch": 0.48,
1170
+ "learning_rate": 0.0002999861225578888,
1171
+ "loss": 0.5407,
1172
+ "step": 171
1173
+ },
1174
+ {
1175
+ "epoch": 0.48,
1176
+ "learning_rate": 0.00029998476945151183,
1177
+ "loss": 0.4942,
1178
+ "step": 172
1179
+ },
1180
+ {
1181
+ "epoch": 0.49,
1182
+ "learning_rate": 0.00029998335341431174,
1183
+ "loss": 0.4563,
1184
+ "step": 173
1185
+ },
1186
+ {
1187
+ "epoch": 0.49,
1188
+ "learning_rate": 0.0002999818744468825,
1189
+ "loss": 0.4406,
1190
+ "step": 174
1191
+ },
1192
+ {
1193
+ "epoch": 0.49,
1194
+ "learning_rate": 0.0002999803325498448,
1195
+ "loss": 0.5319,
1196
+ "step": 175
1197
+ },
1198
+ {
1199
+ "epoch": 0.49,
1200
+ "eval_loss": 0.5312690734863281,
1201
+ "eval_runtime": 26.3302,
1202
+ "eval_samples_per_second": 7.596,
1203
+ "eval_steps_per_second": 1.899,
1204
+ "step": 175
1205
+ },
1206
+ {
1207
+ "dharma_eval_accuracy": 0.40200352864364475,
1208
+ "dharma_eval_accuracy_ARC-Challenge": 0.42592592592592593,
1209
+ "dharma_eval_accuracy_ARC-Easy": 0.5,
1210
+ "dharma_eval_accuracy_BoolQ": 0.7592592592592593,
1211
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
1212
+ "dharma_eval_accuracy_agieval": 0.22033898305084745,
1213
+ "dharma_eval_accuracy_bigbench": 0.36065573770491804,
1214
+ "dharma_eval_accuracy_openbookqa": 0.2222222222222222,
1215
+ "dharma_eval_accuracy_truthful_qa": 0.2962962962962963,
1216
+ "dharma_eval_accuracy_winogrande": 0.42592592592592593,
1217
+ "dharma_loss": 2.2439456572532652,
1218
+ "epoch": 0.49,
1219
+ "step": 175
1220
+ },
1221
+ {
1222
+ "epoch": 0.49,
1223
+ "learning_rate": 0.0002999787277238455,
1224
+ "loss": 0.7389,
1225
+ "step": 176
1226
+ },
1227
+ {
1228
+ "epoch": 0.5,
1229
+ "learning_rate": 0.000299977059969558,
1230
+ "loss": 0.6848,
1231
+ "step": 177
1232
+ },
1233
+ {
1234
+ "epoch": 0.5,
1235
+ "learning_rate": 0.00029997532928768204,
1236
+ "loss": 0.5144,
1237
+ "step": 178
1238
+ },
1239
+ {
1240
+ "epoch": 0.5,
1241
+ "learning_rate": 0.00029997353567894384,
1242
+ "loss": 0.5707,
1243
+ "step": 179
1244
+ },
1245
+ {
1246
+ "epoch": 0.51,
1247
+ "learning_rate": 0.0002999716791440959,
1248
+ "loss": 0.668,
1249
+ "step": 180
1250
+ },
1251
+ {
1252
+ "epoch": 0.51,
1253
+ "learning_rate": 0.00029996975968391715,
1254
+ "loss": 0.4851,
1255
+ "step": 181
1256
+ },
1257
+ {
1258
+ "epoch": 0.51,
1259
+ "learning_rate": 0.000299967777299213,
1260
+ "loss": 0.725,
1261
+ "step": 182
1262
+ },
1263
+ {
1264
+ "epoch": 0.51,
1265
+ "learning_rate": 0.0002999657319908153,
1266
+ "loss": 0.6281,
1267
+ "step": 183
1268
+ },
1269
+ {
1270
+ "epoch": 0.52,
1271
+ "learning_rate": 0.0002999636237595821,
1272
+ "loss": 0.6357,
1273
+ "step": 184
1274
+ },
1275
+ {
1276
+ "epoch": 0.52,
1277
+ "learning_rate": 0.00029996145260639806,
1278
+ "loss": 0.5517,
1279
+ "step": 185
1280
+ },
1281
+ {
1282
+ "epoch": 0.52,
1283
+ "learning_rate": 0.0002999592185321741,
1284
+ "loss": 0.6792,
1285
+ "step": 186
1286
+ },
1287
+ {
1288
+ "epoch": 0.52,
1289
+ "learning_rate": 0.0002999569215378477,
1290
+ "loss": 0.619,
1291
+ "step": 187
1292
+ },
1293
+ {
1294
+ "epoch": 0.53,
1295
+ "learning_rate": 0.0002999545616243825,
1296
+ "loss": 0.5693,
1297
+ "step": 188
1298
+ },
1299
+ {
1300
+ "epoch": 0.53,
1301
+ "learning_rate": 0.00029995213879276876,
1302
+ "loss": 0.5759,
1303
+ "step": 189
1304
+ },
1305
+ {
1306
+ "epoch": 0.53,
1307
+ "learning_rate": 0.000299949653044023,
1308
+ "loss": 0.4382,
1309
+ "step": 190
1310
+ },
1311
+ {
1312
+ "epoch": 0.54,
1313
+ "learning_rate": 0.00029994710437918824,
1314
+ "loss": 0.7183,
1315
+ "step": 191
1316
+ },
1317
+ {
1318
+ "epoch": 0.54,
1319
+ "learning_rate": 0.0002999444927993338,
1320
+ "loss": 0.7328,
1321
+ "step": 192
1322
+ },
1323
+ {
1324
+ "epoch": 0.54,
1325
+ "learning_rate": 0.00029994181830555555,
1326
+ "loss": 0.645,
1327
+ "step": 193
1328
+ },
1329
+ {
1330
+ "epoch": 0.54,
1331
+ "learning_rate": 0.00029993908089897555,
1332
+ "loss": 0.5784,
1333
+ "step": 194
1334
+ },
1335
+ {
1336
+ "epoch": 0.55,
1337
+ "learning_rate": 0.00029993628058074245,
1338
+ "loss": 0.4547,
1339
+ "step": 195
1340
+ },
1341
+ {
1342
+ "epoch": 0.55,
1343
+ "learning_rate": 0.00029993341735203114,
1344
+ "loss": 0.5243,
1345
+ "step": 196
1346
+ },
1347
+ {
1348
+ "epoch": 0.55,
1349
+ "learning_rate": 0.00029993049121404303,
1350
+ "loss": 0.5059,
1351
+ "step": 197
1352
+ },
1353
+ {
1354
+ "epoch": 0.56,
1355
+ "learning_rate": 0.0002999275021680058,
1356
+ "loss": 0.6225,
1357
+ "step": 198
1358
+ },
1359
+ {
1360
+ "epoch": 0.56,
1361
+ "learning_rate": 0.0002999244502151737,
1362
+ "loss": 0.5956,
1363
+ "step": 199
1364
+ },
1365
+ {
1366
+ "epoch": 0.56,
1367
+ "learning_rate": 0.00029992133535682725,
1368
+ "loss": 0.553,
1369
+ "step": 200
1370
+ },
1371
+ {
1372
+ "epoch": 0.56,
1373
+ "eval_loss": 0.5266188979148865,
1374
+ "eval_runtime": 26.3,
1375
+ "eval_samples_per_second": 7.605,
1376
+ "eval_steps_per_second": 1.901,
1377
+ "step": 200
1378
+ },
1379
+ {
1380
+ "dharma_eval_accuracy": 0.4282802607491564,
1381
+ "dharma_eval_accuracy_ARC-Challenge": 0.5185185185185185,
1382
+ "dharma_eval_accuracy_ARC-Easy": 0.5740740740740741,
1383
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1384
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
1385
+ "dharma_eval_accuracy_agieval": 0.22033898305084745,
1386
+ "dharma_eval_accuracy_bigbench": 0.39344262295081966,
1387
+ "dharma_eval_accuracy_openbookqa": 0.14814814814814814,
1388
+ "dharma_eval_accuracy_truthful_qa": 0.3148148148148148,
1389
+ "dharma_eval_accuracy_winogrande": 0.5740740740740741,
1390
+ "dharma_loss": 2.507861177444458,
1391
+ "epoch": 0.56,
1392
+ "step": 200
1393
+ },
1394
+ {
1395
+ "epoch": 0.56,
1396
+ "learning_rate": 0.0002999181575942733,
1397
+ "loss": 0.4542,
1398
+ "step": 201
1399
+ },
1400
+ {
1401
+ "epoch": 0.57,
1402
+ "learning_rate": 0.0002999149169288452,
1403
+ "loss": 0.568,
1404
+ "step": 202
1405
+ },
1406
+ {
1407
+ "epoch": 0.57,
1408
+ "learning_rate": 0.0002999116133619028,
1409
+ "loss": 0.462,
1410
+ "step": 203
1411
+ },
1412
+ {
1413
+ "epoch": 0.57,
1414
+ "learning_rate": 0.0002999082468948321,
1415
+ "loss": 0.4854,
1416
+ "step": 204
1417
+ },
1418
+ {
1419
+ "epoch": 0.58,
1420
+ "learning_rate": 0.00029990481752904563,
1421
+ "loss": 0.474,
1422
+ "step": 205
1423
+ },
1424
+ {
1425
+ "epoch": 0.58,
1426
+ "learning_rate": 0.0002999013252659823,
1427
+ "loss": 0.5331,
1428
+ "step": 206
1429
+ },
1430
+ {
1431
+ "epoch": 0.58,
1432
+ "learning_rate": 0.0002998977701071074,
1433
+ "loss": 0.606,
1434
+ "step": 207
1435
+ },
1436
+ {
1437
+ "epoch": 0.58,
1438
+ "learning_rate": 0.00029989415205391263,
1439
+ "loss": 0.5452,
1440
+ "step": 208
1441
+ },
1442
+ {
1443
+ "epoch": 0.59,
1444
+ "learning_rate": 0.00029989047110791595,
1445
+ "loss": 0.6128,
1446
+ "step": 209
1447
+ },
1448
+ {
1449
+ "epoch": 0.59,
1450
+ "learning_rate": 0.0002998867272706619,
1451
+ "loss": 0.7108,
1452
+ "step": 210
1453
+ },
1454
+ {
1455
+ "epoch": 0.59,
1456
+ "learning_rate": 0.0002998829205437214,
1457
+ "loss": 0.5557,
1458
+ "step": 211
1459
+ },
1460
+ {
1461
+ "epoch": 0.6,
1462
+ "learning_rate": 0.0002998790509286915,
1463
+ "loss": 0.556,
1464
+ "step": 212
1465
+ },
1466
+ {
1467
+ "epoch": 0.6,
1468
+ "learning_rate": 0.000299875118427196,
1469
+ "loss": 0.6482,
1470
+ "step": 213
1471
+ },
1472
+ {
1473
+ "epoch": 0.6,
1474
+ "learning_rate": 0.00029987112304088483,
1475
+ "loss": 0.6108,
1476
+ "step": 214
1477
+ },
1478
+ {
1479
+ "epoch": 0.6,
1480
+ "learning_rate": 0.0002998670647714343,
1481
+ "loss": 0.7099,
1482
+ "step": 215
1483
+ },
1484
+ {
1485
+ "epoch": 0.61,
1486
+ "learning_rate": 0.0002998629436205473,
1487
+ "loss": 0.5939,
1488
+ "step": 216
1489
+ },
1490
+ {
1491
+ "epoch": 0.61,
1492
+ "learning_rate": 0.00029985875958995296,
1493
+ "loss": 0.6238,
1494
+ "step": 217
1495
+ },
1496
+ {
1497
+ "epoch": 0.61,
1498
+ "learning_rate": 0.00029985451268140683,
1499
+ "loss": 0.5144,
1500
+ "step": 218
1501
+ },
1502
+ {
1503
+ "epoch": 0.61,
1504
+ "learning_rate": 0.00029985020289669077,
1505
+ "loss": 0.7351,
1506
+ "step": 219
1507
+ },
1508
+ {
1509
+ "epoch": 0.62,
1510
+ "learning_rate": 0.00029984583023761317,
1511
+ "loss": 0.6136,
1512
+ "step": 220
1513
+ },
1514
+ {
1515
+ "epoch": 0.62,
1516
+ "learning_rate": 0.0002998413947060086,
1517
+ "loss": 0.7544,
1518
+ "step": 221
1519
+ },
1520
+ {
1521
+ "epoch": 0.62,
1522
+ "learning_rate": 0.00029983689630373825,
1523
+ "loss": 0.5588,
1524
+ "step": 222
1525
+ },
1526
+ {
1527
+ "epoch": 0.63,
1528
+ "learning_rate": 0.0002998323350326895,
1529
+ "loss": 0.5725,
1530
+ "step": 223
1531
+ },
1532
+ {
1533
+ "epoch": 0.63,
1534
+ "learning_rate": 0.0002998277108947762,
1535
+ "loss": 0.6262,
1536
+ "step": 224
1537
+ },
1538
+ {
1539
+ "epoch": 0.63,
1540
+ "learning_rate": 0.00029982302389193856,
1541
+ "loss": 0.6217,
1542
+ "step": 225
1543
+ },
1544
+ {
1545
+ "epoch": 0.63,
1546
+ "eval_loss": 0.527147114276886,
1547
+ "eval_runtime": 26.3496,
1548
+ "eval_samples_per_second": 7.59,
1549
+ "eval_steps_per_second": 1.898,
1550
+ "step": 225
1551
+ },
1552
+ {
1553
+ "dharma_eval_accuracy": 0.39483532805751936,
1554
+ "dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
1555
+ "dharma_eval_accuracy_ARC-Easy": 0.6111111111111112,
1556
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1557
+ "dharma_eval_accuracy_MMLU": 0.5185185185185185,
1558
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
1559
+ "dharma_eval_accuracy_bigbench": 0.39344262295081966,
1560
+ "dharma_eval_accuracy_openbookqa": 0.037037037037037035,
1561
+ "dharma_eval_accuracy_truthful_qa": 0.18518518518518517,
1562
+ "dharma_eval_accuracy_winogrande": 0.35185185185185186,
1563
+ "dharma_loss": 2.334738277435303,
1564
+ "epoch": 0.63,
1565
+ "step": 225
1566
+ },
1567
+ {
1568
+ "epoch": 0.63,
1569
+ "learning_rate": 0.0002998182740261432,
1570
+ "loss": 0.6376,
1571
+ "step": 226
1572
+ },
1573
+ {
1574
+ "epoch": 0.64,
1575
+ "learning_rate": 0.0002998134612993829,
1576
+ "loss": 0.5943,
1577
+ "step": 227
1578
+ },
1579
+ {
1580
+ "epoch": 0.64,
1581
+ "learning_rate": 0.0002998085857136772,
1582
+ "loss": 0.4096,
1583
+ "step": 228
1584
+ },
1585
+ {
1586
+ "epoch": 0.64,
1587
+ "learning_rate": 0.00029980364727107166,
1588
+ "loss": 0.5825,
1589
+ "step": 229
1590
+ },
1591
+ {
1592
+ "epoch": 0.65,
1593
+ "learning_rate": 0.00029979864597363845,
1594
+ "loss": 0.5372,
1595
+ "step": 230
1596
+ },
1597
+ {
1598
+ "epoch": 0.65,
1599
+ "learning_rate": 0.0002997935818234759,
1600
+ "loss": 0.5376,
1601
+ "step": 231
1602
+ },
1603
+ {
1604
+ "epoch": 0.65,
1605
+ "learning_rate": 0.00029978845482270906,
1606
+ "loss": 0.4369,
1607
+ "step": 232
1608
+ },
1609
+ {
1610
+ "epoch": 0.65,
1611
+ "learning_rate": 0.0002997832649734889,
1612
+ "loss": 0.5717,
1613
+ "step": 233
1614
+ },
1615
+ {
1616
+ "epoch": 0.66,
1617
+ "learning_rate": 0.0002997780122779931,
1618
+ "loss": 0.6981,
1619
+ "step": 234
1620
+ },
1621
+ {
1622
+ "epoch": 0.66,
1623
+ "learning_rate": 0.0002997726967384255,
1624
+ "loss": 0.5764,
1625
+ "step": 235
1626
+ },
1627
+ {
1628
+ "epoch": 0.66,
1629
+ "learning_rate": 0.0002997673183570165,
1630
+ "loss": 0.5937,
1631
+ "step": 236
1632
+ },
1633
+ {
1634
+ "epoch": 0.67,
1635
+ "learning_rate": 0.00029976187713602273,
1636
+ "loss": 0.2708,
1637
+ "step": 237
1638
+ },
1639
+ {
1640
+ "epoch": 0.67,
1641
+ "learning_rate": 0.0002997563730777273,
1642
+ "loss": 0.6383,
1643
+ "step": 238
1644
+ },
1645
+ {
1646
+ "epoch": 0.67,
1647
+ "learning_rate": 0.00029975080618443946,
1648
+ "loss": 0.491,
1649
+ "step": 239
1650
+ },
1651
+ {
1652
+ "epoch": 0.67,
1653
+ "learning_rate": 0.00029974517645849503,
1654
+ "loss": 0.5049,
1655
+ "step": 240
1656
+ },
1657
+ {
1658
+ "epoch": 0.68,
1659
+ "learning_rate": 0.0002997394839022562,
1660
+ "loss": 0.5768,
1661
+ "step": 241
1662
+ },
1663
+ {
1664
+ "epoch": 0.68,
1665
+ "learning_rate": 0.00029973372851811145,
1666
+ "loss": 0.566,
1667
+ "step": 242
1668
+ },
1669
+ {
1670
+ "epoch": 0.68,
1671
+ "learning_rate": 0.00029972791030847553,
1672
+ "loss": 0.7687,
1673
+ "step": 243
1674
+ },
1675
+ {
1676
+ "epoch": 0.68,
1677
+ "learning_rate": 0.0002997220292757898,
1678
+ "loss": 0.6399,
1679
+ "step": 244
1680
+ },
1681
+ {
1682
+ "epoch": 0.69,
1683
+ "learning_rate": 0.0002997160854225217,
1684
+ "loss": 0.545,
1685
+ "step": 245
1686
+ },
1687
+ {
1688
+ "epoch": 0.69,
1689
+ "learning_rate": 0.00029971007875116527,
1690
+ "loss": 0.6108,
1691
+ "step": 246
1692
+ },
1693
+ {
1694
+ "epoch": 0.69,
1695
+ "learning_rate": 0.0002997040092642407,
1696
+ "loss": 0.6478,
1697
+ "step": 247
1698
+ },
1699
+ {
1700
+ "epoch": 0.7,
1701
+ "learning_rate": 0.0002996978769642947,
1702
+ "loss": 0.5959,
1703
+ "step": 248
1704
+ },
1705
+ {
1706
+ "epoch": 0.7,
1707
+ "learning_rate": 0.0002996916818539003,
1708
+ "loss": 0.5995,
1709
+ "step": 249
1710
+ },
1711
+ {
1712
+ "epoch": 0.7,
1713
+ "learning_rate": 0.0002996854239356567,
1714
+ "loss": 0.5721,
1715
+ "step": 250
1716
+ },
1717
+ {
1718
+ "epoch": 0.7,
1719
+ "eval_loss": 0.5253523588180542,
1720
+ "eval_runtime": 26.3162,
1721
+ "eval_samples_per_second": 7.6,
1722
+ "eval_steps_per_second": 1.9,
1723
+ "step": 250
1724
+ },
1725
+ {
1726
+ "dharma_eval_accuracy": 0.35750614311016893,
1727
+ "dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
1728
+ "dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
1729
+ "dharma_eval_accuracy_BoolQ": 0.6851851851851852,
1730
+ "dharma_eval_accuracy_MMLU": 0.3888888888888889,
1731
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
1732
+ "dharma_eval_accuracy_bigbench": 0.29508196721311475,
1733
+ "dharma_eval_accuracy_openbookqa": 0.018518518518518517,
1734
+ "dharma_eval_accuracy_truthful_qa": 0.2222222222222222,
1735
+ "dharma_eval_accuracy_winogrande": 0.35185185185185186,
1736
+ "dharma_loss": 3.018469988822937,
1737
+ "epoch": 0.7,
1738
+ "step": 250
1739
+ },
1740
+ {
1741
+ "epoch": 0.7,
1742
+ "learning_rate": 0.0002996791032121898,
1743
+ "loss": 0.3632,
1744
+ "step": 251
1745
+ },
1746
+ {
1747
+ "epoch": 0.71,
1748
+ "learning_rate": 0.0002996727196861515,
1749
+ "loss": 0.5306,
1750
+ "step": 252
1751
+ },
1752
+ {
1753
+ "epoch": 0.71,
1754
+ "learning_rate": 0.00029966627336022034,
1755
+ "loss": 0.4332,
1756
+ "step": 253
1757
+ },
1758
+ {
1759
+ "epoch": 0.71,
1760
+ "learning_rate": 0.000299659764237101,
1761
+ "loss": 0.4484,
1762
+ "step": 254
1763
+ },
1764
+ {
1765
+ "epoch": 0.72,
1766
+ "learning_rate": 0.00029965319231952455,
1767
+ "loss": 0.5353,
1768
+ "step": 255
1769
+ },
1770
+ {
1771
+ "epoch": 0.72,
1772
+ "learning_rate": 0.00029964655761024854,
1773
+ "loss": 0.6232,
1774
+ "step": 256
1775
+ },
1776
+ {
1777
+ "epoch": 0.72,
1778
+ "learning_rate": 0.00029963986011205675,
1779
+ "loss": 0.672,
1780
+ "step": 257
1781
+ },
1782
+ {
1783
+ "epoch": 0.72,
1784
+ "learning_rate": 0.0002996330998277593,
1785
+ "loss": 0.5865,
1786
+ "step": 258
1787
+ },
1788
+ {
1789
+ "epoch": 0.73,
1790
+ "learning_rate": 0.0002996262767601926,
1791
+ "loss": 0.8846,
1792
+ "step": 259
1793
+ },
1794
+ {
1795
+ "epoch": 0.73,
1796
+ "learning_rate": 0.0002996193909122197,
1797
+ "loss": 0.4847,
1798
+ "step": 260
1799
+ },
1800
+ {
1801
+ "epoch": 0.73,
1802
+ "learning_rate": 0.00029961244228672953,
1803
+ "loss": 0.5886,
1804
+ "step": 261
1805
+ },
1806
+ {
1807
+ "epoch": 0.74,
1808
+ "learning_rate": 0.0002996054308866378,
1809
+ "loss": 0.6324,
1810
+ "step": 262
1811
+ },
1812
+ {
1813
+ "epoch": 0.74,
1814
+ "learning_rate": 0.0002995983567148862,
1815
+ "loss": 0.4812,
1816
+ "step": 263
1817
+ },
1818
+ {
1819
+ "epoch": 0.74,
1820
+ "learning_rate": 0.000299591219774443,
1821
+ "loss": 0.6549,
1822
+ "step": 264
1823
+ },
1824
+ {
1825
+ "epoch": 0.74,
1826
+ "learning_rate": 0.00029958402006830274,
1827
+ "loss": 0.4712,
1828
+ "step": 265
1829
+ },
1830
+ {
1831
+ "epoch": 0.75,
1832
+ "learning_rate": 0.0002995767575994863,
1833
+ "loss": 0.452,
1834
+ "step": 266
1835
+ },
1836
+ {
1837
+ "epoch": 0.75,
1838
+ "learning_rate": 0.00029956943237104084,
1839
+ "loss": 0.5716,
1840
+ "step": 267
1841
+ },
1842
+ {
1843
+ "epoch": 0.75,
1844
+ "learning_rate": 0.0002995620443860399,
1845
+ "loss": 0.6757,
1846
+ "step": 268
1847
+ },
1848
+ {
1849
+ "epoch": 0.76,
1850
+ "learning_rate": 0.0002995545936475833,
1851
+ "loss": 0.5403,
1852
+ "step": 269
1853
+ },
1854
+ {
1855
+ "epoch": 0.76,
1856
+ "learning_rate": 0.0002995470801587973,
1857
+ "loss": 0.6785,
1858
+ "step": 270
1859
+ },
1860
+ {
1861
+ "epoch": 0.76,
1862
+ "learning_rate": 0.0002995395039228343,
1863
+ "loss": 0.6292,
1864
+ "step": 271
1865
+ },
1866
+ {
1867
+ "epoch": 0.76,
1868
+ "learning_rate": 0.00029953186494287336,
1869
+ "loss": 0.5181,
1870
+ "step": 272
1871
+ },
1872
+ {
1873
+ "epoch": 0.77,
1874
+ "learning_rate": 0.0002995241632221195,
1875
+ "loss": 0.5219,
1876
+ "step": 273
1877
+ },
1878
+ {
1879
+ "epoch": 0.77,
1880
+ "learning_rate": 0.00029951639876380425,
1881
+ "loss": 0.5475,
1882
+ "step": 274
1883
+ },
1884
+ {
1885
+ "epoch": 0.77,
1886
+ "learning_rate": 0.0002995085715711854,
1887
+ "loss": 0.5605,
1888
+ "step": 275
1889
+ },
1890
+ {
1891
+ "epoch": 0.77,
1892
+ "eval_loss": 0.5327661633491516,
1893
+ "eval_runtime": 26.3174,
1894
+ "eval_samples_per_second": 7.6,
1895
+ "eval_steps_per_second": 1.9,
1896
+ "step": 275
1897
+ },
1898
+ {
1899
+ "dharma_eval_accuracy": 0.4329237545408704,
1900
+ "dharma_eval_accuracy_ARC-Challenge": 0.5555555555555556,
1901
+ "dharma_eval_accuracy_ARC-Easy": 0.6111111111111112,
1902
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1903
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
1904
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
1905
+ "dharma_eval_accuracy_bigbench": 0.47540983606557374,
1906
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
1907
+ "dharma_eval_accuracy_truthful_qa": 0.3148148148148148,
1908
+ "dharma_eval_accuracy_winogrande": 0.5185185185185185,
1909
+ "dharma_loss": 2.377507550239563,
1910
+ "epoch": 0.77,
1911
+ "step": 275
1912
+ },
1913
+ {
1914
+ "epoch": 0.77,
1915
+ "learning_rate": 0.0002995006816475471,
1916
+ "loss": 0.5067,
1917
+ "step": 276
1918
+ },
1919
+ {
1920
+ "epoch": 0.78,
1921
+ "learning_rate": 0.00029949272899619994,
1922
+ "loss": 0.6751,
1923
+ "step": 277
1924
+ },
1925
+ {
1926
+ "epoch": 0.78,
1927
+ "learning_rate": 0.0002994847136204805,
1928
+ "loss": 0.5557,
1929
+ "step": 278
1930
+ },
1931
+ {
1932
+ "epoch": 0.78,
1933
+ "learning_rate": 0.0002994766355237521,
1934
+ "loss": 0.6013,
1935
+ "step": 279
1936
+ },
1937
+ {
1938
+ "epoch": 0.79,
1939
+ "learning_rate": 0.0002994684947094039,
1940
+ "loss": 0.4763,
1941
+ "step": 280
1942
+ },
1943
+ {
1944
+ "epoch": 0.79,
1945
+ "learning_rate": 0.00029946029118085193,
1946
+ "loss": 0.5,
1947
+ "step": 281
1948
+ },
1949
+ {
1950
+ "epoch": 0.79,
1951
+ "learning_rate": 0.000299452024941538,
1952
+ "loss": 0.4948,
1953
+ "step": 282
1954
+ },
1955
+ {
1956
+ "epoch": 0.79,
1957
+ "learning_rate": 0.0002994436959949306,
1958
+ "loss": 0.4705,
1959
+ "step": 283
1960
+ },
1961
+ {
1962
+ "epoch": 0.8,
1963
+ "learning_rate": 0.0002994353043445244,
1964
+ "loss": 0.8139,
1965
+ "step": 284
1966
+ },
1967
+ {
1968
+ "epoch": 0.8,
1969
+ "learning_rate": 0.0002994268499938403,
1970
+ "loss": 0.4713,
1971
+ "step": 285
1972
+ },
1973
+ {
1974
+ "epoch": 0.8,
1975
+ "learning_rate": 0.0002994183329464256,
1976
+ "loss": 0.5527,
1977
+ "step": 286
1978
+ },
1979
+ {
1980
+ "epoch": 0.81,
1981
+ "learning_rate": 0.00029940975320585396,
1982
+ "loss": 0.5425,
1983
+ "step": 287
1984
+ },
1985
+ {
1986
+ "epoch": 0.81,
1987
+ "learning_rate": 0.00029940111077572526,
1988
+ "loss": 0.5926,
1989
+ "step": 288
1990
+ },
1991
+ {
1992
+ "epoch": 0.81,
1993
+ "learning_rate": 0.00029939240565966574,
1994
+ "loss": 0.5987,
1995
+ "step": 289
1996
+ },
1997
+ {
1998
+ "epoch": 0.81,
1999
+ "learning_rate": 0.00029938363786132774,
2000
+ "loss": 0.6077,
2001
+ "step": 290
2002
+ },
2003
+ {
2004
+ "epoch": 0.82,
2005
+ "learning_rate": 0.00029937480738439023,
2006
+ "loss": 0.4822,
2007
+ "step": 291
2008
+ },
2009
+ {
2010
+ "epoch": 0.82,
2011
+ "learning_rate": 0.00029936591423255826,
2012
+ "loss": 0.6527,
2013
+ "step": 292
2014
+ },
2015
+ {
2016
+ "epoch": 0.82,
2017
+ "learning_rate": 0.00029935695840956327,
2018
+ "loss": 0.5724,
2019
+ "step": 293
2020
+ },
2021
+ {
2022
+ "epoch": 0.83,
2023
+ "learning_rate": 0.00029934793991916295,
2024
+ "loss": 0.6617,
2025
+ "step": 294
2026
+ },
2027
+ {
2028
+ "epoch": 0.83,
2029
+ "learning_rate": 0.00029933885876514115,
2030
+ "loss": 0.5795,
2031
+ "step": 295
2032
+ },
2033
+ {
2034
+ "epoch": 0.83,
2035
+ "learning_rate": 0.0002993297149513083,
2036
+ "loss": 0.5049,
2037
+ "step": 296
2038
+ },
2039
+ {
2040
+ "epoch": 0.83,
2041
+ "learning_rate": 0.00029932050848150105,
2042
+ "loss": 0.5826,
2043
+ "step": 297
2044
+ },
2045
+ {
2046
+ "epoch": 0.84,
2047
+ "learning_rate": 0.0002993112393595821,
2048
+ "loss": 0.4766,
2049
+ "step": 298
2050
+ },
2051
+ {
2052
+ "epoch": 0.84,
2053
+ "learning_rate": 0.0002993019075894406,
2054
+ "loss": 0.4514,
2055
+ "step": 299
2056
+ },
2057
+ {
2058
+ "epoch": 0.84,
2059
+ "learning_rate": 0.0002992925131749921,
2060
+ "loss": 0.5825,
2061
+ "step": 300
2062
+ },
2063
+ {
2064
+ "epoch": 0.84,
2065
+ "eval_loss": 0.5207599401473999,
2066
+ "eval_runtime": 26.3377,
2067
+ "eval_samples_per_second": 7.594,
2068
+ "eval_steps_per_second": 1.898,
2069
+ "step": 300
2070
+ },
2071
+ {
2072
+ "dharma_eval_accuracy": 0.42874678265681937,
2073
+ "dharma_eval_accuracy_ARC-Challenge": 0.5,
2074
+ "dharma_eval_accuracy_ARC-Easy": 0.6111111111111112,
2075
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2076
+ "dharma_eval_accuracy_MMLU": 0.3333333333333333,
2077
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2078
+ "dharma_eval_accuracy_bigbench": 0.4918032786885246,
2079
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2080
+ "dharma_eval_accuracy_truthful_qa": 0.2962962962962963,
2081
+ "dharma_eval_accuracy_winogrande": 0.5925925925925926,
2082
+ "dharma_loss": 2.452057005882263,
2083
+ "epoch": 0.84,
2084
+ "step": 300
2085
+ },
2086
+ {
2087
+ "epoch": 0.84,
2088
+ "learning_rate": 0.00029928305612017823,
2089
+ "loss": 0.4251,
2090
+ "step": 301
2091
+ },
2092
+ {
2093
+ "epoch": 0.85,
2094
+ "learning_rate": 0.0002992735364289671,
2095
+ "loss": 0.4762,
2096
+ "step": 302
2097
+ },
2098
+ {
2099
+ "epoch": 0.85,
2100
+ "learning_rate": 0.0002992639541053528,
2101
+ "loss": 0.4531,
2102
+ "step": 303
2103
+ },
2104
+ {
2105
+ "epoch": 0.85,
2106
+ "learning_rate": 0.0002992543091533561,
2107
+ "loss": 0.64,
2108
+ "step": 304
2109
+ },
2110
+ {
2111
+ "epoch": 0.86,
2112
+ "learning_rate": 0.00029924460157702376,
2113
+ "loss": 0.6483,
2114
+ "step": 305
2115
+ },
2116
+ {
2117
+ "epoch": 0.86,
2118
+ "learning_rate": 0.0002992348313804289,
2119
+ "loss": 0.5646,
2120
+ "step": 306
2121
+ },
2122
+ {
2123
+ "epoch": 0.86,
2124
+ "learning_rate": 0.00029922499856767094,
2125
+ "loss": 0.4947,
2126
+ "step": 307
2127
+ },
2128
+ {
2129
+ "epoch": 0.86,
2130
+ "learning_rate": 0.00029921510314287545,
2131
+ "loss": 0.4793,
2132
+ "step": 308
2133
+ },
2134
+ {
2135
+ "epoch": 0.87,
2136
+ "learning_rate": 0.00029920514511019456,
2137
+ "loss": 0.3957,
2138
+ "step": 309
2139
+ },
2140
+ {
2141
+ "epoch": 0.87,
2142
+ "learning_rate": 0.00029919512447380625,
2143
+ "loss": 0.6941,
2144
+ "step": 310
2145
+ },
2146
+ {
2147
+ "epoch": 0.87,
2148
+ "learning_rate": 0.0002991850412379151,
2149
+ "loss": 0.6013,
2150
+ "step": 311
2151
+ },
2152
+ {
2153
+ "epoch": 0.88,
2154
+ "learning_rate": 0.0002991748954067519,
2155
+ "loss": 0.6011,
2156
+ "step": 312
2157
+ },
2158
+ {
2159
+ "epoch": 0.88,
2160
+ "learning_rate": 0.0002991646869845736,
2161
+ "loss": 0.7579,
2162
+ "step": 313
2163
+ },
2164
+ {
2165
+ "epoch": 0.88,
2166
+ "learning_rate": 0.0002991544159756634,
2167
+ "loss": 0.4339,
2168
+ "step": 314
2169
+ },
2170
+ {
2171
+ "epoch": 0.88,
2172
+ "learning_rate": 0.0002991440823843309,
2173
+ "loss": 0.4666,
2174
+ "step": 315
2175
+ },
2176
+ {
2177
+ "epoch": 0.89,
2178
+ "learning_rate": 0.0002991336862149119,
2179
+ "loss": 0.5458,
2180
+ "step": 316
2181
+ },
2182
+ {
2183
+ "epoch": 0.89,
2184
+ "learning_rate": 0.00029912322747176835,
2185
+ "loss": 0.6499,
2186
+ "step": 317
2187
+ },
2188
+ {
2189
+ "epoch": 0.89,
2190
+ "learning_rate": 0.0002991127061592887,
2191
+ "loss": 0.7208,
2192
+ "step": 318
2193
+ },
2194
+ {
2195
+ "epoch": 0.9,
2196
+ "learning_rate": 0.00029910212228188734,
2197
+ "loss": 0.7648,
2198
+ "step": 319
2199
+ },
2200
+ {
2201
+ "epoch": 0.9,
2202
+ "learning_rate": 0.0002990914758440051,
2203
+ "loss": 0.9322,
2204
+ "step": 320
2205
+ },
2206
+ {
2207
+ "epoch": 0.9,
2208
+ "learning_rate": 0.00029908076685010915,
2209
+ "loss": 0.6769,
2210
+ "step": 321
2211
+ },
2212
+ {
2213
+ "epoch": 0.9,
2214
+ "learning_rate": 0.0002990699953046926,
2215
+ "loss": 0.6241,
2216
+ "step": 322
2217
+ },
2218
+ {
2219
+ "epoch": 0.91,
2220
+ "learning_rate": 0.00029905916121227515,
2221
+ "loss": 0.409,
2222
+ "step": 323
2223
+ },
2224
+ {
2225
+ "epoch": 0.91,
2226
+ "learning_rate": 0.00029904826457740247,
2227
+ "loss": 0.3138,
2228
+ "step": 324
2229
+ },
2230
+ {
2231
+ "epoch": 0.91,
2232
+ "learning_rate": 0.00029903730540464666,
2233
+ "loss": 0.5241,
2234
+ "step": 325
2235
+ },
2236
+ {
2237
+ "epoch": 0.91,
2238
+ "eval_loss": 0.5201582908630371,
2239
+ "eval_runtime": 26.3949,
2240
+ "eval_samples_per_second": 7.577,
2241
+ "eval_steps_per_second": 1.894,
2242
+ "step": 325
2243
+ },
2244
+ {
2245
+ "dharma_eval_accuracy": 0.3616831149942199,
2246
+ "dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
2247
+ "dharma_eval_accuracy_ARC-Easy": 0.5,
2248
+ "dharma_eval_accuracy_BoolQ": 0.7222222222222222,
2249
+ "dharma_eval_accuracy_MMLU": 0.4444444444444444,
2250
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
2251
+ "dharma_eval_accuracy_bigbench": 0.2786885245901639,
2252
+ "dharma_eval_accuracy_openbookqa": 0.05555555555555555,
2253
+ "dharma_eval_accuracy_truthful_qa": 0.3888888888888889,
2254
+ "dharma_eval_accuracy_winogrande": 0.16666666666666666,
2255
+ "dharma_loss": 2.3925395412445067,
2256
+ "epoch": 0.91,
2257
+ "step": 325
2258
+ },
2259
+ {
2260
+ "epoch": 0.92,
2261
+ "learning_rate": 0.0002990262836986059,
2262
+ "loss": 0.358,
2263
+ "step": 326
2264
+ },
2265
+ {
2266
+ "epoch": 0.92,
2267
+ "learning_rate": 0.0002990151994639048,
2268
+ "loss": 0.5683,
2269
+ "step": 327
2270
+ },
2271
+ {
2272
+ "epoch": 0.92,
2273
+ "learning_rate": 0.000299004052705194,
2274
+ "loss": 0.4744,
2275
+ "step": 328
2276
+ },
2277
+ {
2278
+ "epoch": 0.92,
2279
+ "learning_rate": 0.00029899284342715054,
2280
+ "loss": 0.6101,
2281
+ "step": 329
2282
+ },
2283
+ {
2284
+ "epoch": 0.93,
2285
+ "learning_rate": 0.00029898157163447763,
2286
+ "loss": 0.5381,
2287
+ "step": 330
2288
+ },
2289
+ {
2290
+ "epoch": 0.93,
2291
+ "learning_rate": 0.0002989702373319047,
2292
+ "loss": 0.5269,
2293
+ "step": 331
2294
+ },
2295
+ {
2296
+ "epoch": 0.93,
2297
+ "learning_rate": 0.00029895884052418735,
2298
+ "loss": 0.4993,
2299
+ "step": 332
2300
+ },
2301
+ {
2302
+ "epoch": 0.93,
2303
+ "learning_rate": 0.00029894738121610755,
2304
+ "loss": 0.5623,
2305
+ "step": 333
2306
+ },
2307
+ {
2308
+ "epoch": 0.94,
2309
+ "learning_rate": 0.0002989358594124733,
2310
+ "loss": 0.5988,
2311
+ "step": 334
2312
+ },
2313
+ {
2314
+ "epoch": 0.94,
2315
+ "learning_rate": 0.0002989242751181191,
2316
+ "loss": 0.5623,
2317
+ "step": 335
2318
+ },
2319
+ {
2320
+ "epoch": 0.94,
2321
+ "learning_rate": 0.0002989126283379054,
2322
+ "loss": 0.4471,
2323
+ "step": 336
2324
+ },
2325
+ {
2326
+ "epoch": 0.95,
2327
+ "learning_rate": 0.000298900919076719,
2328
+ "loss": 0.489,
2329
+ "step": 337
2330
+ },
2331
+ {
2332
+ "epoch": 0.95,
2333
+ "learning_rate": 0.00029888914733947275,
2334
+ "loss": 0.5404,
2335
+ "step": 338
2336
+ },
2337
+ {
2338
+ "epoch": 0.95,
2339
+ "learning_rate": 0.00029887731313110613,
2340
+ "loss": 0.5639,
2341
+ "step": 339
2342
+ },
2343
+ {
2344
+ "epoch": 0.95,
2345
+ "learning_rate": 0.0002988654164565843,
2346
+ "loss": 0.6034,
2347
+ "step": 340
2348
+ },
2349
+ {
2350
+ "epoch": 0.96,
2351
+ "learning_rate": 0.00029885345732089905,
2352
+ "loss": 0.5529,
2353
+ "step": 341
2354
+ },
2355
+ {
2356
+ "epoch": 0.96,
2357
+ "learning_rate": 0.0002988414357290681,
2358
+ "loss": 0.6584,
2359
+ "step": 342
2360
+ },
2361
+ {
2362
+ "epoch": 0.96,
2363
+ "learning_rate": 0.0002988293516861356,
2364
+ "loss": 0.5119,
2365
+ "step": 343
2366
+ },
2367
+ {
2368
+ "epoch": 0.97,
2369
+ "learning_rate": 0.0002988172051971717,
2370
+ "loss": 0.4797,
2371
+ "step": 344
2372
+ },
2373
+ {
2374
+ "epoch": 0.97,
2375
+ "learning_rate": 0.0002988049962672728,
2376
+ "loss": 0.619,
2377
+ "step": 345
2378
+ },
2379
+ {
2380
+ "epoch": 0.97,
2381
+ "learning_rate": 0.0002987927249015616,
2382
+ "loss": 0.621,
2383
+ "step": 346
2384
+ },
2385
+ {
2386
+ "epoch": 0.97,
2387
+ "learning_rate": 0.00029878039110518704,
2388
+ "loss": 0.6019,
2389
+ "step": 347
2390
+ },
2391
+ {
2392
+ "epoch": 0.98,
2393
+ "learning_rate": 0.000298767994883324,
2394
+ "loss": 0.5099,
2395
+ "step": 348
2396
+ },
2397
+ {
2398
+ "epoch": 0.98,
2399
+ "learning_rate": 0.00029875553624117375,
2400
+ "loss": 0.3973,
2401
+ "step": 349
2402
+ },
2403
+ {
2404
+ "epoch": 0.98,
2405
+ "learning_rate": 0.00029874301518396376,
2406
+ "loss": 0.6068,
2407
+ "step": 350
2408
+ },
2409
+ {
2410
+ "epoch": 0.98,
2411
+ "eval_loss": 0.517635703086853,
2412
+ "eval_runtime": 26.3545,
2413
+ "eval_samples_per_second": 7.589,
2414
+ "eval_steps_per_second": 1.897,
2415
+ "step": 350
2416
+ },
2417
+ {
2418
+ "dharma_eval_accuracy": 0.36985182212251455,
2419
+ "dharma_eval_accuracy_ARC-Challenge": 0.37037037037037035,
2420
+ "dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
2421
+ "dharma_eval_accuracy_BoolQ": 0.6851851851851852,
2422
+ "dharma_eval_accuracy_MMLU": 0.2962962962962963,
2423
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2424
+ "dharma_eval_accuracy_bigbench": 0.29508196721311475,
2425
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2426
+ "dharma_eval_accuracy_truthful_qa": 0.35185185185185186,
2427
+ "dharma_eval_accuracy_winogrande": 0.5555555555555556,
2428
+ "dharma_loss": 2.8203083801269533,
2429
+ "epoch": 0.98,
2430
+ "step": 350
2431
+ },
2432
+ {
2433
+ "epoch": 0.99,
2434
+ "learning_rate": 0.00029873043171694755,
2435
+ "loss": 0.5346,
2436
+ "step": 351
2437
+ },
2438
+ {
2439
+ "epoch": 0.99,
2440
+ "learning_rate": 0.0002987177858454049,
2441
+ "loss": 0.3792,
2442
+ "step": 352
2443
+ },
2444
+ {
2445
+ "epoch": 0.99,
2446
+ "learning_rate": 0.00029870507757464193,
2447
+ "loss": 0.5227,
2448
+ "step": 353
2449
+ },
2450
+ {
2451
+ "epoch": 0.99,
2452
+ "learning_rate": 0.0002986923069099906,
2453
+ "loss": 0.4218,
2454
+ "step": 354
2455
+ },
2456
+ {
2457
+ "epoch": 1.0,
2458
+ "learning_rate": 0.00029867947385680936,
2459
+ "loss": 0.4866,
2460
+ "step": 355
2461
+ },
2462
+ {
2463
+ "epoch": 1.0,
2464
+ "learning_rate": 0.00029866657842048274,
2465
+ "loss": 0.4593,
2466
+ "step": 356
2467
+ },
2468
+ {
2469
+ "epoch": 1.0,
2470
+ "learning_rate": 0.00029865362060642136,
2471
+ "loss": 0.6367,
2472
+ "step": 357
2473
+ },
2474
+ {
2475
+ "epoch": 1.0,
2476
+ "learning_rate": 0.0002986406004200621,
2477
+ "loss": 0.5064,
2478
+ "step": 358
2479
+ },
2480
+ {
2481
+ "epoch": 1.01,
2482
+ "learning_rate": 0.00029862751786686797,
2483
+ "loss": 0.4206,
2484
+ "step": 359
2485
+ },
2486
+ {
2487
+ "epoch": 1.01,
2488
+ "learning_rate": 0.0002986143729523282,
2489
+ "loss": 0.6043,
2490
+ "step": 360
2491
+ },
2492
+ {
2493
+ "epoch": 1.01,
2494
+ "learning_rate": 0.0002986011656819582,
2495
+ "loss": 0.6889,
2496
+ "step": 361
2497
+ },
2498
+ {
2499
+ "epoch": 1.02,
2500
+ "learning_rate": 0.0002985878960612993,
2501
+ "loss": 0.6266,
2502
+ "step": 362
2503
+ },
2504
+ {
2505
+ "epoch": 1.02,
2506
+ "learning_rate": 0.0002985745640959194,
2507
+ "loss": 0.6016,
2508
+ "step": 363
2509
+ },
2510
+ {
2511
+ "epoch": 1.02,
2512
+ "learning_rate": 0.00029856116979141224,
2513
+ "loss": 0.387,
2514
+ "step": 364
2515
+ },
2516
+ {
2517
+ "epoch": 1.02,
2518
+ "learning_rate": 0.00029854771315339785,
2519
+ "loss": 0.4954,
2520
+ "step": 365
2521
+ },
2522
+ {
2523
+ "epoch": 1.03,
2524
+ "learning_rate": 0.0002985341941875224,
2525
+ "loss": 0.5165,
2526
+ "step": 366
2527
+ },
2528
+ {
2529
+ "epoch": 1.03,
2530
+ "learning_rate": 0.0002985206128994581,
2531
+ "loss": 0.459,
2532
+ "step": 367
2533
+ },
2534
+ {
2535
+ "epoch": 1.03,
2536
+ "learning_rate": 0.0002985069692949036,
2537
+ "loss": 0.5966,
2538
+ "step": 368
2539
+ },
2540
+ {
2541
+ "epoch": 1.04,
2542
+ "learning_rate": 0.0002984932633795833,
2543
+ "loss": 0.4271,
2544
+ "step": 369
2545
+ },
2546
+ {
2547
+ "epoch": 1.04,
2548
+ "learning_rate": 0.00029847949515924806,
2549
+ "loss": 0.4243,
2550
+ "step": 370
2551
+ },
2552
+ {
2553
+ "epoch": 1.04,
2554
+ "learning_rate": 0.00029846566463967477,
2555
+ "loss": 0.4879,
2556
+ "step": 371
2557
+ },
2558
+ {
2559
+ "epoch": 1.04,
2560
+ "learning_rate": 0.0002984517718266664,
2561
+ "loss": 0.4225,
2562
+ "step": 372
2563
+ },
2564
+ {
2565
+ "epoch": 1.05,
2566
+ "learning_rate": 0.00029843781672605216,
2567
+ "loss": 0.385,
2568
+ "step": 373
2569
+ },
2570
+ {
2571
+ "epoch": 1.05,
2572
+ "learning_rate": 0.00029842379934368735,
2573
+ "loss": 0.5161,
2574
+ "step": 374
2575
+ },
2576
+ {
2577
+ "epoch": 1.05,
2578
+ "learning_rate": 0.0002984097196854534,
2579
+ "loss": 0.5034,
2580
+ "step": 375
2581
+ },
2582
+ {
2583
+ "epoch": 1.05,
2584
+ "eval_loss": 0.5201339721679688,
2585
+ "eval_runtime": 26.3832,
2586
+ "eval_samples_per_second": 7.581,
2587
+ "eval_steps_per_second": 1.895,
2588
+ "step": 375
2589
+ },
2590
+ {
2591
+ "dharma_eval_accuracy": 0.40587691825690037,
2592
+ "dharma_eval_accuracy_ARC-Challenge": 0.24074074074074073,
2593
+ "dharma_eval_accuracy_ARC-Easy": 0.2962962962962963,
2594
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2595
+ "dharma_eval_accuracy_MMLU": 0.2037037037037037,
2596
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2597
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2598
+ "dharma_eval_accuracy_openbookqa": 0.24074074074074073,
2599
+ "dharma_eval_accuracy_truthful_qa": 0.7037037037037037,
2600
+ "dharma_eval_accuracy_winogrande": 0.5185185185185185,
2601
+ "dharma_loss": 2.7163280713558198,
2602
+ "epoch": 1.05,
2603
+ "step": 375
2604
+ },
2605
+ {
2606
+ "epoch": 1.06,
2607
+ "learning_rate": 0.0002983955777572578,
2608
+ "loss": 0.6131,
2609
+ "step": 376
2610
+ },
2611
+ {
2612
+ "epoch": 1.06,
2613
+ "learning_rate": 0.0002983813735650344,
2614
+ "loss": 0.546,
2615
+ "step": 377
2616
+ },
2617
+ {
2618
+ "epoch": 1.06,
2619
+ "learning_rate": 0.00029836710711474287,
2620
+ "loss": 0.3439,
2621
+ "step": 378
2622
+ },
2623
+ {
2624
+ "epoch": 1.06,
2625
+ "learning_rate": 0.0002983527784123692,
2626
+ "loss": 0.6843,
2627
+ "step": 379
2628
+ },
2629
+ {
2630
+ "epoch": 1.07,
2631
+ "learning_rate": 0.0002983383874639254,
2632
+ "loss": 0.4989,
2633
+ "step": 380
2634
+ },
2635
+ {
2636
+ "epoch": 1.07,
2637
+ "learning_rate": 0.0002983239342754498,
2638
+ "loss": 0.4311,
2639
+ "step": 381
2640
+ },
2641
+ {
2642
+ "epoch": 1.07,
2643
+ "learning_rate": 0.0002983094188530065,
2644
+ "loss": 0.6035,
2645
+ "step": 382
2646
+ },
2647
+ {
2648
+ "epoch": 1.08,
2649
+ "learning_rate": 0.000298294841202686,
2650
+ "loss": 0.3606,
2651
+ "step": 383
2652
+ },
2653
+ {
2654
+ "epoch": 1.08,
2655
+ "learning_rate": 0.0002982802013306048,
2656
+ "loss": 0.3556,
2657
+ "step": 384
2658
+ },
2659
+ {
2660
+ "epoch": 1.08,
2661
+ "learning_rate": 0.00029826549924290557,
2662
+ "loss": 0.4458,
2663
+ "step": 385
2664
+ },
2665
+ {
2666
+ "epoch": 1.08,
2667
+ "learning_rate": 0.0002982507349457569,
2668
+ "loss": 0.4737,
2669
+ "step": 386
2670
+ },
2671
+ {
2672
+ "epoch": 1.09,
2673
+ "learning_rate": 0.00029823590844535366,
2674
+ "loss": 0.5703,
2675
+ "step": 387
2676
+ },
2677
+ {
2678
+ "epoch": 1.09,
2679
+ "learning_rate": 0.0002982210197479169,
2680
+ "loss": 0.3601,
2681
+ "step": 388
2682
+ },
2683
+ {
2684
+ "epoch": 1.09,
2685
+ "learning_rate": 0.00029820606885969347,
2686
+ "loss": 0.337,
2687
+ "step": 389
2688
+ },
2689
+ {
2690
+ "epoch": 1.09,
2691
+ "learning_rate": 0.00029819105578695655,
2692
+ "loss": 0.4757,
2693
+ "step": 390
2694
+ },
2695
+ {
2696
+ "epoch": 1.1,
2697
+ "learning_rate": 0.0002981759805360054,
2698
+ "loss": 0.5072,
2699
+ "step": 391
2700
+ },
2701
+ {
2702
+ "epoch": 1.1,
2703
+ "learning_rate": 0.0002981608431131653,
2704
+ "loss": 0.4409,
2705
+ "step": 392
2706
+ },
2707
+ {
2708
+ "epoch": 1.1,
2709
+ "learning_rate": 0.00029814564352478753,
2710
+ "loss": 0.5252,
2711
+ "step": 393
2712
+ },
2713
+ {
2714
+ "epoch": 1.11,
2715
+ "learning_rate": 0.00029813038177724965,
2716
+ "loss": 0.5016,
2717
+ "step": 394
2718
+ },
2719
+ {
2720
+ "epoch": 1.11,
2721
+ "learning_rate": 0.00029811505787695524,
2722
+ "loss": 0.3186,
2723
+ "step": 395
2724
+ },
2725
+ {
2726
+ "epoch": 1.11,
2727
+ "learning_rate": 0.0002980996718303338,
2728
+ "loss": 0.5822,
2729
+ "step": 396
2730
+ },
2731
+ {
2732
+ "epoch": 1.11,
2733
+ "learning_rate": 0.00029808422364384113,
2734
+ "loss": 0.54,
2735
+ "step": 397
2736
+ },
2737
+ {
2738
+ "epoch": 1.12,
2739
+ "learning_rate": 0.00029806871332395895,
2740
+ "loss": 0.4582,
2741
+ "step": 398
2742
+ },
2743
+ {
2744
+ "epoch": 1.12,
2745
+ "learning_rate": 0.0002980531408771951,
2746
+ "loss": 0.5073,
2747
+ "step": 399
2748
+ },
2749
+ {
2750
+ "epoch": 1.12,
2751
+ "learning_rate": 0.00029803750631008356,
2752
+ "loss": 0.5227,
2753
+ "step": 400
2754
+ },
2755
+ {
2756
+ "epoch": 1.12,
2757
+ "eval_loss": 0.5233827829360962,
2758
+ "eval_runtime": 26.3434,
2759
+ "eval_samples_per_second": 7.592,
2760
+ "eval_steps_per_second": 1.898,
2761
+ "step": 400
2762
+ },
2763
+ {
2764
+ "dharma_eval_accuracy": 0.3770703338947604,
2765
+ "dharma_eval_accuracy_ARC-Challenge": 0.24074074074074073,
2766
+ "dharma_eval_accuracy_ARC-Easy": 0.3148148148148148,
2767
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2768
+ "dharma_eval_accuracy_MMLU": 0.2037037037037037,
2769
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2770
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2771
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2772
+ "dharma_eval_accuracy_truthful_qa": 0.5555555555555556,
2773
+ "dharma_eval_accuracy_winogrande": 0.5370370370370371,
2774
+ "dharma_loss": 2.5418553066253664,
2775
+ "epoch": 1.12,
2776
+ "step": 400
2777
+ },
2778
+ {
2779
+ "epoch": 1.13,
2780
+ "learning_rate": 0.00029802180962918426,
2781
+ "loss": 0.3211,
2782
+ "step": 401
2783
+ },
2784
+ {
2785
+ "epoch": 1.13,
2786
+ "learning_rate": 0.00029800605084108315,
2787
+ "loss": 0.5834,
2788
+ "step": 402
2789
+ },
2790
+ {
2791
+ "epoch": 1.13,
2792
+ "learning_rate": 0.0002979902299523925,
2793
+ "loss": 0.6354,
2794
+ "step": 403
2795
+ },
2796
+ {
2797
+ "epoch": 1.13,
2798
+ "learning_rate": 0.00029797434696975035,
2799
+ "loss": 0.4923,
2800
+ "step": 404
2801
+ },
2802
+ {
2803
+ "epoch": 1.14,
2804
+ "learning_rate": 0.0002979584018998209,
2805
+ "loss": 0.4609,
2806
+ "step": 405
2807
+ },
2808
+ {
2809
+ "epoch": 1.14,
2810
+ "learning_rate": 0.0002979423947492944,
2811
+ "loss": 0.4805,
2812
+ "step": 406
2813
+ },
2814
+ {
2815
+ "epoch": 1.14,
2816
+ "learning_rate": 0.0002979263255248872,
2817
+ "loss": 0.4223,
2818
+ "step": 407
2819
+ },
2820
+ {
2821
+ "epoch": 1.15,
2822
+ "learning_rate": 0.0002979101942333416,
2823
+ "loss": 0.4872,
2824
+ "step": 408
2825
+ },
2826
+ {
2827
+ "epoch": 1.15,
2828
+ "learning_rate": 0.00029789400088142605,
2829
+ "loss": 0.4437,
2830
+ "step": 409
2831
+ },
2832
+ {
2833
+ "epoch": 1.15,
2834
+ "learning_rate": 0.000297877745475935,
2835
+ "loss": 0.7126,
2836
+ "step": 410
2837
+ },
2838
+ {
2839
+ "epoch": 1.15,
2840
+ "learning_rate": 0.00029786142802368877,
2841
+ "loss": 0.3855,
2842
+ "step": 411
2843
+ },
2844
+ {
2845
+ "epoch": 1.16,
2846
+ "learning_rate": 0.00029784504853153397,
2847
+ "loss": 0.4374,
2848
+ "step": 412
2849
+ },
2850
+ {
2851
+ "epoch": 1.16,
2852
+ "learning_rate": 0.0002978286070063431,
2853
+ "loss": 0.4274,
2854
+ "step": 413
2855
+ },
2856
+ {
2857
+ "epoch": 1.16,
2858
+ "learning_rate": 0.0002978121034550148,
2859
+ "loss": 0.406,
2860
+ "step": 414
2861
+ },
2862
+ {
2863
+ "epoch": 1.16,
2864
+ "learning_rate": 0.00029779553788447357,
2865
+ "loss": 0.4755,
2866
+ "step": 415
2867
+ },
2868
+ {
2869
+ "epoch": 1.17,
2870
+ "learning_rate": 0.00029777891030167,
2871
+ "loss": 0.5286,
2872
+ "step": 416
2873
+ },
2874
+ {
2875
+ "epoch": 1.17,
2876
+ "learning_rate": 0.00029776222071358074,
2877
+ "loss": 0.6529,
2878
+ "step": 417
2879
+ },
2880
+ {
2881
+ "epoch": 1.17,
2882
+ "learning_rate": 0.0002977454691272084,
2883
+ "loss": 0.4096,
2884
+ "step": 418
2885
+ },
2886
+ {
2887
+ "epoch": 1.18,
2888
+ "learning_rate": 0.0002977286555495818,
2889
+ "loss": 0.4292,
2890
+ "step": 419
2891
+ },
2892
+ {
2893
+ "epoch": 1.18,
2894
+ "learning_rate": 0.00029771177998775536,
2895
+ "loss": 0.4458,
2896
+ "step": 420
2897
+ },
2898
+ {
2899
+ "epoch": 1.18,
2900
+ "learning_rate": 0.0002976948424488099,
2901
+ "loss": 0.5428,
2902
+ "step": 421
2903
+ },
2904
+ {
2905
+ "epoch": 1.18,
2906
+ "learning_rate": 0.0002976778429398521,
2907
+ "loss": 0.5953,
2908
+ "step": 422
2909
+ },
2910
+ {
2911
+ "epoch": 1.19,
2912
+ "learning_rate": 0.00029766078146801453,
2913
+ "loss": 0.3686,
2914
+ "step": 423
2915
+ },
2916
+ {
2917
+ "epoch": 1.19,
2918
+ "learning_rate": 0.000297643658040456,
2919
+ "loss": 0.6322,
2920
+ "step": 424
2921
+ },
2922
+ {
2923
+ "epoch": 1.19,
2924
+ "learning_rate": 0.0002976264726643611,
2925
+ "loss": 0.35,
2926
+ "step": 425
2927
+ },
2928
+ {
2929
+ "epoch": 1.19,
2930
+ "eval_loss": 0.5209073424339294,
2931
+ "eval_runtime": 26.4295,
2932
+ "eval_samples_per_second": 7.567,
2933
+ "eval_steps_per_second": 1.892,
2934
+ "step": 425
2935
+ },
2936
+ {
2937
+ "dharma_eval_accuracy": 0.3832431734009332,
2938
+ "dharma_eval_accuracy_ARC-Challenge": 0.25925925925925924,
2939
+ "dharma_eval_accuracy_ARC-Easy": 0.35185185185185186,
2940
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2941
+ "dharma_eval_accuracy_MMLU": 0.2222222222222222,
2942
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2943
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2944
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
2945
+ "dharma_eval_accuracy_truthful_qa": 0.5370370370370371,
2946
+ "dharma_eval_accuracy_winogrande": 0.5185185185185185,
2947
+ "dharma_loss": 2.4492057392597197,
2948
+ "epoch": 1.19,
2949
+ "step": 425
2950
+ },
2951
+ {
2952
+ "epoch": 1.2,
2953
+ "learning_rate": 0.00029760922534694055,
2954
+ "loss": 0.5357,
2955
+ "step": 426
2956
+ },
2957
+ {
2958
+ "epoch": 1.2,
2959
+ "learning_rate": 0.00029759191609543095,
2960
+ "loss": 0.5112,
2961
+ "step": 427
2962
+ },
2963
+ {
2964
+ "epoch": 1.2,
2965
+ "learning_rate": 0.0002975745449170949,
2966
+ "loss": 0.3245,
2967
+ "step": 428
2968
+ },
2969
+ {
2970
+ "epoch": 1.2,
2971
+ "learning_rate": 0.0002975571118192212,
2972
+ "loss": 0.6413,
2973
+ "step": 429
2974
+ },
2975
+ {
2976
+ "epoch": 1.21,
2977
+ "learning_rate": 0.00029753961680912427,
2978
+ "loss": 0.8456,
2979
+ "step": 430
2980
+ },
2981
+ {
2982
+ "epoch": 1.21,
2983
+ "learning_rate": 0.00029752205989414475,
2984
+ "loss": 0.5009,
2985
+ "step": 431
2986
+ },
2987
+ {
2988
+ "epoch": 1.21,
2989
+ "learning_rate": 0.0002975044410816492,
2990
+ "loss": 0.6083,
2991
+ "step": 432
2992
+ },
2993
+ {
2994
+ "epoch": 1.22,
2995
+ "learning_rate": 0.0002974867603790302,
2996
+ "loss": 0.5866,
2997
+ "step": 433
2998
+ },
2999
+ {
3000
+ "epoch": 1.22,
3001
+ "learning_rate": 0.0002974690177937062,
3002
+ "loss": 0.5901,
3003
+ "step": 434
3004
+ },
3005
+ {
3006
+ "epoch": 1.22,
3007
+ "learning_rate": 0.00029745121333312163,
3008
+ "loss": 0.4596,
3009
+ "step": 435
3010
+ },
3011
+ {
3012
+ "epoch": 1.22,
3013
+ "learning_rate": 0.00029743334700474693,
3014
+ "loss": 0.4795,
3015
+ "step": 436
3016
+ },
3017
+ {
3018
+ "epoch": 1.23,
3019
+ "learning_rate": 0.00029741541881607854,
3020
+ "loss": 0.5234,
3021
+ "step": 437
3022
+ },
3023
+ {
3024
+ "epoch": 1.23,
3025
+ "learning_rate": 0.00029739742877463865,
3026
+ "loss": 0.4583,
3027
+ "step": 438
3028
+ },
3029
+ {
3030
+ "epoch": 1.23,
3031
+ "learning_rate": 0.0002973793768879757,
3032
+ "loss": 0.4084,
3033
+ "step": 439
3034
+ },
3035
+ {
3036
+ "epoch": 1.24,
3037
+ "learning_rate": 0.0002973612631636638,
3038
+ "loss": 0.5177,
3039
+ "step": 440
3040
+ },
3041
+ {
3042
+ "epoch": 1.24,
3043
+ "learning_rate": 0.0002973430876093033,
3044
+ "loss": 0.5485,
3045
+ "step": 441
3046
+ },
3047
+ {
3048
+ "epoch": 1.24,
3049
+ "learning_rate": 0.00029732485023252015,
3050
+ "loss": 0.7071,
3051
+ "step": 442
3052
+ },
3053
+ {
3054
+ "epoch": 1.24,
3055
+ "learning_rate": 0.0002973065510409665,
3056
+ "loss": 0.4424,
3057
+ "step": 443
3058
+ },
3059
+ {
3060
+ "epoch": 1.25,
3061
+ "learning_rate": 0.00029728819004232036,
3062
+ "loss": 0.5394,
3063
+ "step": 444
3064
+ },
3065
+ {
3066
+ "epoch": 1.25,
3067
+ "learning_rate": 0.0002972697672442856,
3068
+ "loss": 0.4934,
3069
+ "step": 445
3070
+ },
3071
+ {
3072
+ "epoch": 1.25,
3073
+ "learning_rate": 0.0002972512826545922,
3074
+ "loss": 0.5253,
3075
+ "step": 446
3076
+ },
3077
+ {
3078
+ "epoch": 1.25,
3079
+ "learning_rate": 0.0002972327362809958,
3080
+ "loss": 0.414,
3081
+ "step": 447
3082
+ },
3083
+ {
3084
+ "epoch": 1.26,
3085
+ "learning_rate": 0.0002972141281312782,
3086
+ "loss": 0.3736,
3087
+ "step": 448
3088
+ },
3089
+ {
3090
+ "epoch": 1.26,
3091
+ "learning_rate": 0.000297195458213247,
3092
+ "loss": 0.471,
3093
+ "step": 449
3094
+ },
3095
+ {
3096
+ "epoch": 1.26,
3097
+ "learning_rate": 0.0002971767265347358,
3098
+ "loss": 0.5088,
3099
+ "step": 450
3100
+ },
3101
+ {
3102
+ "epoch": 1.26,
3103
+ "eval_loss": 0.5178927779197693,
3104
+ "eval_runtime": 26.409,
3105
+ "eval_samples_per_second": 7.573,
3106
+ "eval_steps_per_second": 1.893,
3107
+ "step": 450
3108
+ },
3109
+ {
3110
+ "dharma_eval_accuracy": 0.38306879940358374,
3111
+ "dharma_eval_accuracy_ARC-Challenge": 0.25925925925925924,
3112
+ "dharma_eval_accuracy_ARC-Easy": 0.35185185185185186,
3113
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3114
+ "dharma_eval_accuracy_MMLU": 0.2777777777777778,
3115
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3116
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3117
+ "dharma_eval_accuracy_openbookqa": 0.18518518518518517,
3118
+ "dharma_eval_accuracy_truthful_qa": 0.46296296296296297,
3119
+ "dharma_eval_accuracy_winogrande": 0.4444444444444444,
3120
+ "dharma_loss": 2.727772394180298,
3121
+ "epoch": 1.26,
3122
+ "step": 450
3123
+ },
3124
+ {
3125
+ "epoch": 1.27,
3126
+ "learning_rate": 0.0002971579331036041,
3127
+ "loss": 0.5458,
3128
+ "step": 451
3129
+ },
3130
+ {
3131
+ "epoch": 1.27,
3132
+ "learning_rate": 0.00029713907792773716,
3133
+ "loss": 0.6182,
3134
+ "step": 452
3135
+ },
3136
+ {
3137
+ "epoch": 1.27,
3138
+ "learning_rate": 0.0002971201610150463,
3139
+ "loss": 0.2314,
3140
+ "step": 453
3141
+ },
3142
+ {
3143
+ "epoch": 1.27,
3144
+ "learning_rate": 0.0002971011823734688,
3145
+ "loss": 0.479,
3146
+ "step": 454
3147
+ },
3148
+ {
3149
+ "epoch": 1.28,
3150
+ "learning_rate": 0.00029708214201096755,
3151
+ "loss": 0.3984,
3152
+ "step": 455
3153
+ },
3154
+ {
3155
+ "epoch": 1.28,
3156
+ "learning_rate": 0.0002970630399355317,
3157
+ "loss": 0.542,
3158
+ "step": 456
3159
+ },
3160
+ {
3161
+ "epoch": 1.28,
3162
+ "learning_rate": 0.00029704387615517606,
3163
+ "loss": 0.3715,
3164
+ "step": 457
3165
+ },
3166
+ {
3167
+ "epoch": 1.29,
3168
+ "learning_rate": 0.00029702465067794144,
3169
+ "loss": 0.4325,
3170
+ "step": 458
3171
+ },
3172
+ {
3173
+ "epoch": 1.29,
3174
+ "learning_rate": 0.00029700536351189445,
3175
+ "loss": 0.5227,
3176
+ "step": 459
3177
+ },
3178
+ {
3179
+ "epoch": 1.29,
3180
+ "learning_rate": 0.0002969860146651276,
3181
+ "loss": 0.6047,
3182
+ "step": 460
3183
+ },
3184
+ {
3185
+ "epoch": 1.29,
3186
+ "learning_rate": 0.0002969666041457594,
3187
+ "loss": 0.598,
3188
+ "step": 461
3189
+ },
3190
+ {
3191
+ "epoch": 1.3,
3192
+ "learning_rate": 0.00029694713196193404,
3193
+ "loss": 0.5142,
3194
+ "step": 462
3195
+ },
3196
+ {
3197
+ "epoch": 1.3,
3198
+ "learning_rate": 0.0002969275981218218,
3199
+ "loss": 0.3851,
3200
+ "step": 463
3201
+ },
3202
+ {
3203
+ "epoch": 1.3,
3204
+ "learning_rate": 0.0002969080026336186,
3205
+ "loss": 0.5747,
3206
+ "step": 464
3207
+ },
3208
+ {
3209
+ "epoch": 1.31,
3210
+ "learning_rate": 0.00029688834550554646,
3211
+ "loss": 0.4373,
3212
+ "step": 465
3213
+ },
3214
+ {
3215
+ "epoch": 1.31,
3216
+ "learning_rate": 0.00029686862674585307,
3217
+ "loss": 0.3868,
3218
+ "step": 466
3219
+ },
3220
+ {
3221
+ "epoch": 1.31,
3222
+ "learning_rate": 0.00029684884636281203,
3223
+ "loss": 0.6756,
3224
+ "step": 467
3225
+ },
3226
+ {
3227
+ "epoch": 1.31,
3228
+ "learning_rate": 0.00029682900436472286,
3229
+ "loss": 0.5035,
3230
+ "step": 468
3231
+ },
3232
+ {
3233
+ "epoch": 1.32,
3234
+ "learning_rate": 0.00029680910075991087,
3235
+ "loss": 0.4708,
3236
+ "step": 469
3237
+ },
3238
+ {
3239
+ "epoch": 1.32,
3240
+ "learning_rate": 0.0002967891355567273,
3241
+ "loss": 0.5669,
3242
+ "step": 470
3243
+ },
3244
+ {
3245
+ "epoch": 1.32,
3246
+ "learning_rate": 0.0002967691087635491,
3247
+ "loss": 0.4667,
3248
+ "step": 471
3249
+ },
3250
+ {
3251
+ "epoch": 1.32,
3252
+ "learning_rate": 0.0002967490203887793,
3253
+ "loss": 0.5232,
3254
+ "step": 472
3255
+ },
3256
+ {
3257
+ "epoch": 1.33,
3258
+ "learning_rate": 0.00029672887044084636,
3259
+ "loss": 0.408,
3260
+ "step": 473
3261
+ },
3262
+ {
3263
+ "epoch": 1.33,
3264
+ "learning_rate": 0.000296708658928205,
3265
+ "loss": 0.3419,
3266
+ "step": 474
3267
+ },
3268
+ {
3269
+ "epoch": 1.33,
3270
+ "learning_rate": 0.00029668838585933556,
3271
+ "loss": 0.4646,
3272
+ "step": 475
3273
+ },
3274
+ {
3275
+ "epoch": 1.33,
3276
+ "eval_loss": 0.5185211300849915,
3277
+ "eval_runtime": 26.3942,
3278
+ "eval_samples_per_second": 7.577,
3279
+ "eval_steps_per_second": 1.894,
3280
+ "step": 475
3281
+ },
3282
+ {
3283
+ "dharma_eval_accuracy": 0.39335686524720515,
3284
+ "dharma_eval_accuracy_ARC-Challenge": 0.2962962962962963,
3285
+ "dharma_eval_accuracy_ARC-Easy": 0.3333333333333333,
3286
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3287
+ "dharma_eval_accuracy_MMLU": 0.25925925925925924,
3288
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3289
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3290
+ "dharma_eval_accuracy_openbookqa": 0.16666666666666666,
3291
+ "dharma_eval_accuracy_truthful_qa": 0.5185185185185185,
3292
+ "dharma_eval_accuracy_winogrande": 0.5,
3293
+ "dharma_loss": 2.565536605358124,
3294
+ "epoch": 1.33,
3295
+ "step": 475
3296
+ },
3297
+ {
3298
+ "epoch": 1.34,
3299
+ "learning_rate": 0.00029666805124274425,
3300
+ "loss": 0.3971,
3301
+ "step": 476
3302
+ },
3303
+ {
3304
+ "epoch": 1.34,
3305
+ "learning_rate": 0.00029664765508696306,
3306
+ "loss": 0.4487,
3307
+ "step": 477
3308
+ },
3309
+ {
3310
+ "epoch": 1.34,
3311
+ "learning_rate": 0.0002966271974005499,
3312
+ "loss": 0.5002,
3313
+ "step": 478
3314
+ },
3315
+ {
3316
+ "epoch": 1.34,
3317
+ "learning_rate": 0.00029660667819208836,
3318
+ "loss": 0.5285,
3319
+ "step": 479
3320
+ },
3321
+ {
3322
+ "epoch": 1.35,
3323
+ "learning_rate": 0.0002965860974701879,
3324
+ "loss": 0.5782,
3325
+ "step": 480
3326
+ },
3327
+ {
3328
+ "epoch": 1.35,
3329
+ "learning_rate": 0.00029656545524348396,
3330
+ "loss": 0.5334,
3331
+ "step": 481
3332
+ },
3333
+ {
3334
+ "epoch": 1.35,
3335
+ "learning_rate": 0.0002965447515206375,
3336
+ "loss": 0.2758,
3337
+ "step": 482
3338
+ },
3339
+ {
3340
+ "epoch": 1.36,
3341
+ "learning_rate": 0.00029652398631033547,
3342
+ "loss": 0.3964,
3343
+ "step": 483
3344
+ },
3345
+ {
3346
+ "epoch": 1.36,
3347
+ "learning_rate": 0.0002965031596212905,
3348
+ "loss": 0.5087,
3349
+ "step": 484
3350
+ },
3351
+ {
3352
+ "epoch": 1.36,
3353
+ "learning_rate": 0.0002964822714622412,
3354
+ "loss": 0.6133,
3355
+ "step": 485
3356
+ },
3357
+ {
3358
+ "epoch": 1.36,
3359
+ "learning_rate": 0.0002964613218419517,
3360
+ "loss": 0.3947,
3361
+ "step": 486
3362
+ },
3363
+ {
3364
+ "epoch": 1.37,
3365
+ "learning_rate": 0.0002964403107692122,
3366
+ "loss": 0.4799,
3367
+ "step": 487
3368
+ },
3369
+ {
3370
+ "epoch": 1.37,
3371
+ "learning_rate": 0.00029641923825283854,
3372
+ "loss": 0.5146,
3373
+ "step": 488
3374
+ },
3375
+ {
3376
+ "epoch": 1.37,
3377
+ "learning_rate": 0.0002963981043016723,
3378
+ "loss": 0.348,
3379
+ "step": 489
3380
+ },
3381
+ {
3382
+ "epoch": 1.38,
3383
+ "learning_rate": 0.000296376908924581,
3384
+ "loss": 0.5181,
3385
+ "step": 490
3386
+ },
3387
+ {
3388
+ "epoch": 1.38,
3389
+ "learning_rate": 0.0002963556521304577,
3390
+ "loss": 0.4437,
3391
+ "step": 491
3392
+ },
3393
+ {
3394
+ "epoch": 1.38,
3395
+ "learning_rate": 0.0002963343339282214,
3396
+ "loss": 0.5235,
3397
+ "step": 492
3398
+ },
3399
+ {
3400
+ "epoch": 1.38,
3401
+ "learning_rate": 0.0002963129543268168,
3402
+ "loss": 0.3413,
3403
+ "step": 493
3404
+ },
3405
+ {
3406
+ "epoch": 1.39,
3407
+ "learning_rate": 0.00029629151333521446,
3408
+ "loss": 0.4321,
3409
+ "step": 494
3410
+ },
3411
+ {
3412
+ "epoch": 1.39,
3413
+ "learning_rate": 0.00029627001096241057,
3414
+ "loss": 0.3405,
3415
+ "step": 495
3416
+ },
3417
+ {
3418
+ "epoch": 1.39,
3419
+ "learning_rate": 0.0002962484472174271,
3420
+ "loss": 0.4895,
3421
+ "step": 496
3422
+ },
3423
+ {
3424
+ "epoch": 1.4,
3425
+ "learning_rate": 0.0002962268221093118,
3426
+ "loss": 0.4285,
3427
+ "step": 497
3428
+ },
3429
+ {
3430
+ "epoch": 1.4,
3431
+ "learning_rate": 0.0002962051356471383,
3432
+ "loss": 0.4785,
3433
+ "step": 498
3434
+ },
3435
+ {
3436
+ "epoch": 1.4,
3437
+ "learning_rate": 0.0002961833878400056,
3438
+ "loss": 0.618,
3439
+ "step": 499
3440
+ },
3441
+ {
3442
+ "epoch": 1.4,
3443
+ "learning_rate": 0.0002961615786970389,
3444
+ "loss": 0.663,
3445
+ "step": 500
3446
+ },
3447
+ {
3448
+ "epoch": 1.4,
3449
+ "eval_loss": 0.5207934379577637,
3450
+ "eval_runtime": 26.4199,
3451
+ "eval_samples_per_second": 7.57,
3452
+ "eval_steps_per_second": 1.893,
3453
+ "step": 500
3454
+ },
3455
+ {
3456
+ "dharma_eval_accuracy": 0.39952970475337796,
3457
+ "dharma_eval_accuracy_ARC-Challenge": 0.3148148148148148,
3458
+ "dharma_eval_accuracy_ARC-Easy": 0.3888888888888889,
3459
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3460
+ "dharma_eval_accuracy_MMLU": 0.3148148148148148,
3461
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3462
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3463
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
3464
+ "dharma_eval_accuracy_truthful_qa": 0.5185185185185185,
3465
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
3466
+ "dharma_loss": 2.4424794154167175,
3467
+ "epoch": 1.4,
3468
+ "step": 500
3469
+ },
3470
+ {
3471
+ "epoch": 1.41,
3472
+ "learning_rate": 0.00029613970822738874,
3473
+ "loss": 0.365,
3474
+ "step": 501
3475
+ },
3476
+ {
3477
+ "epoch": 1.41,
3478
+ "learning_rate": 0.0002961177764402317,
3479
+ "loss": 0.6666,
3480
+ "step": 502
3481
+ },
3482
+ {
3483
+ "epoch": 1.41,
3484
+ "learning_rate": 0.00029609578334476987,
3485
+ "loss": 0.4039,
3486
+ "step": 503
3487
+ },
3488
+ {
3489
+ "epoch": 1.41,
3490
+ "learning_rate": 0.0002960737289502311,
3491
+ "loss": 0.3414,
3492
+ "step": 504
3493
+ },
3494
+ {
3495
+ "epoch": 1.42,
3496
+ "learning_rate": 0.00029605161326586916,
3497
+ "loss": 0.7406,
3498
+ "step": 505
3499
+ },
3500
+ {
3501
+ "epoch": 1.42,
3502
+ "learning_rate": 0.00029602943630096325,
3503
+ "loss": 0.5311,
3504
+ "step": 506
3505
+ },
3506
+ {
3507
+ "epoch": 1.42,
3508
+ "learning_rate": 0.00029600719806481844,
3509
+ "loss": 0.5669,
3510
+ "step": 507
3511
+ },
3512
+ {
3513
+ "epoch": 1.43,
3514
+ "learning_rate": 0.0002959848985667655,
3515
+ "loss": 0.6251,
3516
+ "step": 508
3517
+ },
3518
+ {
3519
+ "epoch": 1.43,
3520
+ "learning_rate": 0.00029596253781616084,
3521
+ "loss": 0.5971,
3522
+ "step": 509
3523
+ },
3524
+ {
3525
+ "epoch": 1.43,
3526
+ "learning_rate": 0.0002959401158223867,
3527
+ "loss": 0.6097,
3528
+ "step": 510
3529
+ },
3530
+ {
3531
+ "epoch": 1.43,
3532
+ "learning_rate": 0.00029591763259485083,
3533
+ "loss": 0.7195,
3534
+ "step": 511
3535
+ },
3536
+ {
3537
+ "epoch": 1.44,
3538
+ "learning_rate": 0.0002958950881429869,
3539
+ "loss": 0.344,
3540
+ "step": 512
3541
+ },
3542
+ {
3543
+ "epoch": 1.44,
3544
+ "learning_rate": 0.000295872482476254,
3545
+ "loss": 0.5158,
3546
+ "step": 513
3547
+ },
3548
+ {
3549
+ "epoch": 1.44,
3550
+ "learning_rate": 0.00029584981560413717,
3551
+ "loss": 0.4253,
3552
+ "step": 514
3553
+ },
3554
+ {
3555
+ "epoch": 1.45,
3556
+ "learning_rate": 0.0002958270875361469,
3557
+ "loss": 0.5837,
3558
+ "step": 515
3559
+ },
3560
+ {
3561
+ "epoch": 1.45,
3562
+ "learning_rate": 0.0002958042982818196,
3563
+ "loss": 0.4485,
3564
+ "step": 516
3565
+ },
3566
+ {
3567
+ "epoch": 1.45,
3568
+ "learning_rate": 0.0002957814478507171,
3569
+ "loss": 0.2953,
3570
+ "step": 517
3571
+ },
3572
+ {
3573
+ "epoch": 1.45,
3574
+ "learning_rate": 0.00029575853625242704,
3575
+ "loss": 0.446,
3576
+ "step": 518
3577
+ },
3578
+ {
3579
+ "epoch": 1.46,
3580
+ "learning_rate": 0.00029573556349656277,
3581
+ "loss": 0.4378,
3582
+ "step": 519
3583
+ },
3584
+ {
3585
+ "epoch": 1.46,
3586
+ "learning_rate": 0.0002957125295927631,
3587
+ "loss": 0.4399,
3588
+ "step": 520
3589
+ },
3590
+ {
3591
+ "epoch": 1.46,
3592
+ "learning_rate": 0.00029568943455069276,
3593
+ "loss": 0.4011,
3594
+ "step": 521
3595
+ },
3596
+ {
3597
+ "epoch": 1.47,
3598
+ "learning_rate": 0.00029566627838004193,
3599
+ "loss": 0.5164,
3600
+ "step": 522
3601
+ },
3602
+ {
3603
+ "epoch": 1.47,
3604
+ "learning_rate": 0.0002956430610905265,
3605
+ "loss": 0.454,
3606
+ "step": 523
3607
+ },
3608
+ {
3609
+ "epoch": 1.47,
3610
+ "learning_rate": 0.00029561978269188814,
3611
+ "loss": 0.6102,
3612
+ "step": 524
3613
+ },
3614
+ {
3615
+ "epoch": 1.47,
3616
+ "learning_rate": 0.0002955964431938939,
3617
+ "loss": 0.4802,
3618
+ "step": 525
3619
+ },
3620
+ {
3621
+ "epoch": 1.47,
3622
+ "eval_loss": 0.5176565051078796,
3623
+ "eval_runtime": 26.4201,
3624
+ "eval_samples_per_second": 7.57,
3625
+ "eval_steps_per_second": 1.892,
3626
+ "step": 525
3627
+ },
3628
+ {
3629
+ "dharma_eval_accuracy": 0.41669896873502815,
3630
+ "dharma_eval_accuracy_ARC-Challenge": 0.35185185185185186,
3631
+ "dharma_eval_accuracy_ARC-Easy": 0.48148148148148145,
3632
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3633
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
3634
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3635
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
3636
+ "dharma_eval_accuracy_openbookqa": 0.12962962962962962,
3637
+ "dharma_eval_accuracy_truthful_qa": 0.5185185185185185,
3638
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
3639
+ "dharma_loss": 2.5137911858558657,
3640
+ "epoch": 1.47,
3641
+ "step": 525
3642
+ },
3643
+ {
3644
+ "epoch": 1.48,
3645
+ "learning_rate": 0.0002955730426063365,
3646
+ "loss": 0.364,
3647
+ "step": 526
3648
+ },
3649
+ {
3650
+ "epoch": 1.48,
3651
+ "learning_rate": 0.00029554958093903466,
3652
+ "loss": 0.3682,
3653
+ "step": 527
3654
+ },
3655
+ {
3656
+ "epoch": 1.48,
3657
+ "learning_rate": 0.00029552605820183235,
3658
+ "loss": 0.3183,
3659
+ "step": 528
3660
+ },
3661
+ {
3662
+ "epoch": 1.48,
3663
+ "learning_rate": 0.00029550247440459916,
3664
+ "loss": 0.5107,
3665
+ "step": 529
3666
+ },
3667
+ {
3668
+ "epoch": 1.49,
3669
+ "learning_rate": 0.0002954788295572305,
3670
+ "loss": 0.3863,
3671
+ "step": 530
3672
+ },
3673
+ {
3674
+ "epoch": 1.49,
3675
+ "learning_rate": 0.00029545512366964735,
3676
+ "loss": 0.5892,
3677
+ "step": 531
3678
+ },
3679
+ {
3680
+ "epoch": 1.49,
3681
+ "learning_rate": 0.00029543135675179626,
3682
+ "loss": 0.5724,
3683
+ "step": 532
3684
+ },
3685
+ {
3686
+ "epoch": 1.5,
3687
+ "learning_rate": 0.00029540752881364925,
3688
+ "loss": 0.5255,
3689
+ "step": 533
3690
+ },
3691
+ {
3692
+ "epoch": 1.5,
3693
+ "learning_rate": 0.00029538363986520425,
3694
+ "loss": 0.6092,
3695
+ "step": 534
3696
+ },
3697
+ {
3698
+ "epoch": 1.5,
3699
+ "learning_rate": 0.00029535968991648454,
3700
+ "loss": 0.4652,
3701
+ "step": 535
3702
+ },
3703
+ {
3704
+ "epoch": 1.5,
3705
+ "learning_rate": 0.00029533567897753905,
3706
+ "loss": 0.4963,
3707
+ "step": 536
3708
+ },
3709
+ {
3710
+ "epoch": 1.51,
3711
+ "learning_rate": 0.0002953116070584424,
3712
+ "loss": 0.4462,
3713
+ "step": 537
3714
+ },
3715
+ {
3716
+ "epoch": 1.51,
3717
+ "learning_rate": 0.00029528747416929463,
3718
+ "loss": 0.599,
3719
+ "step": 538
3720
+ },
3721
+ {
3722
+ "epoch": 1.51,
3723
+ "learning_rate": 0.00029526328032022155,
3724
+ "loss": 0.6611,
3725
+ "step": 539
3726
+ },
3727
+ {
3728
+ "epoch": 1.52,
3729
+ "learning_rate": 0.00029523902552137433,
3730
+ "loss": 0.5107,
3731
+ "step": 540
3732
+ },
3733
+ {
3734
+ "epoch": 1.52,
3735
+ "learning_rate": 0.00029521470978292994,
3736
+ "loss": 0.4466,
3737
+ "step": 541
3738
+ },
3739
+ {
3740
+ "epoch": 1.52,
3741
+ "learning_rate": 0.00029519033311509077,
3742
+ "loss": 0.3413,
3743
+ "step": 542
3744
+ },
3745
+ {
3746
+ "epoch": 1.52,
3747
+ "learning_rate": 0.0002951658955280848,
3748
+ "loss": 0.3824,
3749
+ "step": 543
3750
+ },
3751
+ {
3752
+ "epoch": 1.53,
3753
+ "learning_rate": 0.0002951413970321657,
3754
+ "loss": 0.4518,
3755
+ "step": 544
3756
+ },
3757
+ {
3758
+ "epoch": 1.53,
3759
+ "learning_rate": 0.0002951168376376124,
3760
+ "loss": 0.4747,
3761
+ "step": 545
3762
+ },
3763
+ {
3764
+ "epoch": 1.53,
3765
+ "learning_rate": 0.0002950922173547296,
3766
+ "loss": 0.526,
3767
+ "step": 546
3768
+ },
3769
+ {
3770
+ "epoch": 1.54,
3771
+ "learning_rate": 0.00029506753619384766,
3772
+ "loss": 0.571,
3773
+ "step": 547
3774
+ },
3775
+ {
3776
+ "epoch": 1.54,
3777
+ "learning_rate": 0.00029504279416532223,
3778
+ "loss": 0.5026,
3779
+ "step": 548
3780
+ },
3781
+ {
3782
+ "epoch": 1.54,
3783
+ "learning_rate": 0.00029501799127953465,
3784
+ "loss": 0.5395,
3785
+ "step": 549
3786
+ },
3787
+ {
3788
+ "epoch": 1.54,
3789
+ "learning_rate": 0.0002949931275468917,
3790
+ "loss": 0.2888,
3791
+ "step": 550
3792
+ },
3793
+ {
3794
+ "epoch": 1.54,
3795
+ "eval_loss": 0.521288275718689,
3796
+ "eval_runtime": 26.3822,
3797
+ "eval_samples_per_second": 7.581,
3798
+ "eval_steps_per_second": 1.895,
3799
+ "step": 550
3800
+ },
3801
+ {
3802
+ "dharma_eval_accuracy": 0.40364493109082655,
3803
+ "dharma_eval_accuracy_ARC-Challenge": 0.3148148148148148,
3804
+ "dharma_eval_accuracy_ARC-Easy": 0.46296296296296297,
3805
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3806
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
3807
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3808
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3809
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
3810
+ "dharma_eval_accuracy_truthful_qa": 0.4444444444444444,
3811
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
3812
+ "dharma_loss": 2.6077233877182007,
3813
+ "epoch": 1.54,
3814
+ "step": 550
3815
+ },
3816
+ {
3817
+ "epoch": 1.55,
3818
+ "learning_rate": 0.00029496820297782574,
3819
+ "loss": 0.5497,
3820
+ "step": 551
3821
+ },
3822
+ {
3823
+ "epoch": 1.55,
3824
+ "learning_rate": 0.00029494321758279465,
3825
+ "loss": 0.4195,
3826
+ "step": 552
3827
+ },
3828
+ {
3829
+ "epoch": 1.55,
3830
+ "learning_rate": 0.0002949181713722819,
3831
+ "loss": 0.5639,
3832
+ "step": 553
3833
+ },
3834
+ {
3835
+ "epoch": 1.56,
3836
+ "learning_rate": 0.0002948930643567964,
3837
+ "loss": 0.5605,
3838
+ "step": 554
3839
+ },
3840
+ {
3841
+ "epoch": 1.56,
3842
+ "learning_rate": 0.00029486789654687253,
3843
+ "loss": 0.5815,
3844
+ "step": 555
3845
+ },
3846
+ {
3847
+ "epoch": 1.56,
3848
+ "learning_rate": 0.00029484266795307037,
3849
+ "loss": 0.4015,
3850
+ "step": 556
3851
+ },
3852
+ {
3853
+ "epoch": 1.56,
3854
+ "learning_rate": 0.0002948173785859752,
3855
+ "loss": 0.4135,
3856
+ "step": 557
3857
+ },
3858
+ {
3859
+ "epoch": 1.57,
3860
+ "learning_rate": 0.000294792028456198,
3861
+ "loss": 0.4335,
3862
+ "step": 558
3863
+ },
3864
+ {
3865
+ "epoch": 1.57,
3866
+ "learning_rate": 0.0002947666175743753,
3867
+ "loss": 0.512,
3868
+ "step": 559
3869
+ },
3870
+ {
3871
+ "epoch": 1.57,
3872
+ "learning_rate": 0.00029474114595116896,
3873
+ "loss": 0.4886,
3874
+ "step": 560
3875
+ },
3876
+ {
3877
+ "epoch": 1.57,
3878
+ "learning_rate": 0.00029471561359726645,
3879
+ "loss": 0.4351,
3880
+ "step": 561
3881
+ },
3882
+ {
3883
+ "epoch": 1.58,
3884
+ "learning_rate": 0.0002946900205233807,
3885
+ "loss": 0.4052,
3886
+ "step": 562
3887
+ },
3888
+ {
3889
+ "epoch": 1.58,
3890
+ "learning_rate": 0.00029466436674024997,
3891
+ "loss": 0.4267,
3892
+ "step": 563
3893
+ },
3894
+ {
3895
+ "epoch": 1.58,
3896
+ "learning_rate": 0.0002946386522586382,
3897
+ "loss": 0.5737,
3898
+ "step": 564
3899
+ },
3900
+ {
3901
+ "epoch": 1.59,
3902
+ "learning_rate": 0.00029461287708933473,
3903
+ "loss": 0.2997,
3904
+ "step": 565
3905
+ },
3906
+ {
3907
+ "epoch": 1.59,
3908
+ "learning_rate": 0.00029458704124315425,
3909
+ "loss": 0.523,
3910
+ "step": 566
3911
+ },
3912
+ {
3913
+ "epoch": 1.59,
3914
+ "learning_rate": 0.0002945611447309371,
3915
+ "loss": 0.4795,
3916
+ "step": 567
3917
+ },
3918
+ {
3919
+ "epoch": 1.59,
3920
+ "learning_rate": 0.00029453518756354885,
3921
+ "loss": 0.5125,
3922
+ "step": 568
3923
+ },
3924
+ {
3925
+ "epoch": 1.6,
3926
+ "learning_rate": 0.0002945091697518808,
3927
+ "loss": 0.4387,
3928
+ "step": 569
3929
+ },
3930
+ {
3931
+ "epoch": 1.6,
3932
+ "learning_rate": 0.0002944830913068494,
3933
+ "loss": 0.6114,
3934
+ "step": 570
3935
+ },
3936
+ {
3937
+ "epoch": 1.6,
3938
+ "learning_rate": 0.0002944569522393968,
3939
+ "loss": 0.3893,
3940
+ "step": 571
3941
+ },
3942
+ {
3943
+ "epoch": 1.61,
3944
+ "learning_rate": 0.00029443075256049036,
3945
+ "loss": 0.6212,
3946
+ "step": 572
3947
+ },
3948
+ {
3949
+ "epoch": 1.61,
3950
+ "learning_rate": 0.0002944044922811231,
3951
+ "loss": 0.333,
3952
+ "step": 573
3953
+ },
3954
+ {
3955
+ "epoch": 1.61,
3956
+ "learning_rate": 0.0002943781714123132,
3957
+ "loss": 0.4745,
3958
+ "step": 574
3959
+ },
3960
+ {
3961
+ "epoch": 1.61,
3962
+ "learning_rate": 0.00029435178996510455,
3963
+ "loss": 0.5085,
3964
+ "step": 575
3965
+ },
3966
+ {
3967
+ "epoch": 1.61,
3968
+ "eval_loss": 0.5148590207099915,
3969
+ "eval_runtime": 26.387,
3970
+ "eval_samples_per_second": 7.58,
3971
+ "eval_steps_per_second": 1.895,
3972
+ "step": 575
3973
+ },
3974
+ {
3975
+ "dharma_eval_accuracy": 0.42216344960934504,
3976
+ "dharma_eval_accuracy_ARC-Challenge": 0.3148148148148148,
3977
+ "dharma_eval_accuracy_ARC-Easy": 0.37037037037037035,
3978
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3979
+ "dharma_eval_accuracy_MMLU": 0.35185185185185186,
3980
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3981
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3982
+ "dharma_eval_accuracy_openbookqa": 0.18518518518518517,
3983
+ "dharma_eval_accuracy_truthful_qa": 0.5925925925925926,
3984
+ "dharma_eval_accuracy_winogrande": 0.5185185185185185,
3985
+ "dharma_loss": 2.7804353098869323,
3986
+ "epoch": 1.61,
3987
+ "step": 575
3988
+ },
3989
+ {
3990
+ "epoch": 1.62,
3991
+ "learning_rate": 0.0002943253479505662,
3992
+ "loss": 0.4255,
3993
+ "step": 576
3994
+ },
3995
+ {
3996
+ "epoch": 1.62,
3997
+ "learning_rate": 0.0002942988453797928,
3998
+ "loss": 0.4885,
3999
+ "step": 577
4000
+ },
4001
+ {
4002
+ "epoch": 1.62,
4003
+ "learning_rate": 0.00029427228226390424,
4004
+ "loss": 0.5474,
4005
+ "step": 578
4006
+ },
4007
+ {
4008
+ "epoch": 1.63,
4009
+ "learning_rate": 0.00029424565861404606,
4010
+ "loss": 0.5646,
4011
+ "step": 579
4012
+ },
4013
+ {
4014
+ "epoch": 1.63,
4015
+ "learning_rate": 0.00029421897444138897,
4016
+ "loss": 0.3399,
4017
+ "step": 580
4018
+ },
4019
+ {
4020
+ "epoch": 1.63,
4021
+ "learning_rate": 0.00029419222975712915,
4022
+ "loss": 0.4264,
4023
+ "step": 581
4024
+ },
4025
+ {
4026
+ "epoch": 1.63,
4027
+ "learning_rate": 0.00029416542457248816,
4028
+ "loss": 0.5138,
4029
+ "step": 582
4030
+ },
4031
+ {
4032
+ "epoch": 1.64,
4033
+ "learning_rate": 0.000294138558898713,
4034
+ "loss": 0.5768,
4035
+ "step": 583
4036
+ },
4037
+ {
4038
+ "epoch": 1.64,
4039
+ "learning_rate": 0.000294111632747076,
4040
+ "loss": 0.475,
4041
+ "step": 584
4042
+ },
4043
+ {
4044
+ "epoch": 1.64,
4045
+ "learning_rate": 0.0002940846461288748,
4046
+ "loss": 0.4983,
4047
+ "step": 585
4048
+ },
4049
+ {
4050
+ "epoch": 1.64,
4051
+ "learning_rate": 0.00029405759905543256,
4052
+ "loss": 0.4168,
4053
+ "step": 586
4054
+ },
4055
+ {
4056
+ "epoch": 1.65,
4057
+ "learning_rate": 0.00029403049153809774,
4058
+ "loss": 0.6149,
4059
+ "step": 587
4060
+ },
4061
+ {
4062
+ "epoch": 1.65,
4063
+ "learning_rate": 0.0002940033235882441,
4064
+ "loss": 0.4848,
4065
+ "step": 588
4066
+ },
4067
+ {
4068
+ "epoch": 1.65,
4069
+ "learning_rate": 0.0002939760952172708,
4070
+ "loss": 0.5926,
4071
+ "step": 589
4072
+ },
4073
+ {
4074
+ "epoch": 1.66,
4075
+ "learning_rate": 0.0002939488064366024,
4076
+ "loss": 0.4797,
4077
+ "step": 590
4078
+ },
4079
+ {
4080
+ "epoch": 1.66,
4081
+ "learning_rate": 0.00029392145725768874,
4082
+ "loss": 0.5197,
4083
+ "step": 591
4084
+ },
4085
+ {
4086
+ "epoch": 1.66,
4087
+ "learning_rate": 0.0002938940476920051,
4088
+ "loss": 0.4958,
4089
+ "step": 592
4090
+ },
4091
+ {
4092
+ "epoch": 1.66,
4093
+ "learning_rate": 0.0002938665777510519,
4094
+ "loss": 0.7065,
4095
+ "step": 593
4096
+ },
4097
+ {
4098
+ "epoch": 1.67,
4099
+ "learning_rate": 0.0002938390474463551,
4100
+ "loss": 0.4155,
4101
+ "step": 594
4102
+ },
4103
+ {
4104
+ "epoch": 1.67,
4105
+ "learning_rate": 0.00029381145678946584,
4106
+ "loss": 0.522,
4107
+ "step": 595
4108
+ },
4109
+ {
4110
+ "epoch": 1.67,
4111
+ "learning_rate": 0.00029378380579196076,
4112
+ "loss": 0.4253,
4113
+ "step": 596
4114
+ },
4115
+ {
4116
+ "epoch": 1.68,
4117
+ "learning_rate": 0.00029375609446544165,
4118
+ "loss": 0.6296,
4119
+ "step": 597
4120
+ },
4121
+ {
4122
+ "epoch": 1.68,
4123
+ "learning_rate": 0.0002937283228215356,
4124
+ "loss": 0.515,
4125
+ "step": 598
4126
+ },
4127
+ {
4128
+ "epoch": 1.68,
4129
+ "learning_rate": 0.00029370049087189514,
4130
+ "loss": 0.6629,
4131
+ "step": 599
4132
+ },
4133
+ {
4134
+ "epoch": 1.68,
4135
+ "learning_rate": 0.00029367259862819804,
4136
+ "loss": 0.5582,
4137
+ "step": 600
4138
+ },
4139
+ {
4140
+ "epoch": 1.68,
4141
+ "eval_loss": 0.510773241519928,
4142
+ "eval_runtime": 26.4135,
4143
+ "eval_samples_per_second": 7.572,
4144
+ "eval_steps_per_second": 1.893,
4145
+ "step": 600
4146
+ },
4147
+ {
4148
+ "dharma_eval_accuracy": 0.4201058364406208,
4149
+ "dharma_eval_accuracy_ARC-Challenge": 0.2962962962962963,
4150
+ "dharma_eval_accuracy_ARC-Easy": 0.3888888888888889,
4151
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
4152
+ "dharma_eval_accuracy_MMLU": 0.2962962962962963,
4153
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
4154
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
4155
+ "dharma_eval_accuracy_openbookqa": 0.24074074074074073,
4156
+ "dharma_eval_accuracy_truthful_qa": 0.5740740740740741,
4157
+ "dharma_eval_accuracy_winogrande": 0.5185185185185185,
4158
+ "dharma_loss": 2.6925288438796997,
4159
+ "epoch": 1.68,
4160
+ "step": 600
4161
+ }
4162
+ ],
4163
+ "max_steps": 5000,
4164
+ "num_train_epochs": 15,
4165
+ "total_flos": 1.9307426648459674e+17,
4166
+ "trial_name": null,
4167
+ "trial_params": null
4168
+ }
checkpoint-600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0e686fbe23782f1d8fdfb6f62df77c86b1440a419242d0d49479cf2bd76fa6f
3
+ size 5947