Hugging Face's logo

Peter
/
qloratry

PEFT
Model card Files
xet
Community
1
Peter commited on
Commit
64eefb5
·
1 Parent(s): 4dfa28f

Upload folder using huggingface_hub

Browse files
Files changed (18) hide show
  1. README.md +21 -0
  2. adapter_config.json +21 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-400/README.md +34 -0
  5. checkpoint-400/adapter_config.json +21 -0
  6. checkpoint-400/adapter_model.bin +3 -0
  7. checkpoint-400/optimizer.pt +3 -0
  8. checkpoint-400/rng_state.pth +3 -0
  9. checkpoint-400/scheduler.pt +3 -0
  10. checkpoint-400/special_tokens_map.json +24 -0
  11. checkpoint-400/tokenizer.json +0 -0
  12. checkpoint-400/tokenizer_config.json +34 -0
  13. checkpoint-400/trainer_state.json +2419 -0
  14. checkpoint-400/training_args.bin +3 -0
  15. special_tokens_map.json +24 -0
  16. tokenizer.json +0 -0
  17. tokenizer_config.json +34 -0
  18. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/Llama-2-70b-hf",
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,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f061286d3e4ee5b0a3b5005f04980859a9ee00997a98d2ac99e4282457307b3
3
+ size 524404298
checkpoint-400/README.md ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+
19
+ The following `bitsandbytes` quantization config was used during training:
20
+ - quant_method: bitsandbytes
21
+ - load_in_8bit: False
22
+ - load_in_4bit: True
23
+ - llm_int8_threshold: 6.0
24
+ - llm_int8_skip_modules: None
25
+ - llm_int8_enable_fp32_cpu_offload: False
26
+ - llm_int8_has_fp16_weight: False
27
+ - bnb_4bit_quant_type: nf4
28
+ - bnb_4bit_use_double_quant: True
29
+ - bnb_4bit_compute_dtype: bfloat16
30
+ ### Framework versions
31
+
32
+ - PEFT 0.5.0
33
+
34
+ - PEFT 0.5.0
checkpoint-400/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/Llama-2-70b-hf",
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,
11
+ "lora_dropout": 0.0,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-400/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a5ef628f0d7895f294cb7f374da0a9029f6dbd767d609f44170885323522a89
3
+ size 524404298
checkpoint-400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c0d895362ab0fe23a3418cd5f68ee0e0a5880c2469385aa86a388c722e997ce
3
+ size 263038356
checkpoint-400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c062f7f375beded48b5337f5a3f3a5cb38807fa3e85dbf3e294c0ab6b627bfc2
3
+ size 14244
checkpoint-400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c20cbcf7b82efaf4fff9d5357d8e31a4d82ca6a38f1ffe38178b73623aeef25d
3
+ size 1064
checkpoint-400/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
checkpoint-400/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-400/tokenizer_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "__type": "AddedToken",
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "clean_up_tokenization_spaces": false,
11
+ "eos_token": {
12
+ "__type": "AddedToken",
13
+ "content": "</s>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false
18
+ },
19
+ "legacy": false,
20
+ "model_max_length": 1000000000000000019884624838656,
21
+ "pad_token": null,
22
+ "padding_side": "right",
23
+ "sp_model_kwargs": {},
24
+ "tokenizer_class": "LlamaTokenizer",
25
+ "unk_token": {
26
+ "__type": "AddedToken",
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ },
33
+ "use_default_system_prompt": true
34
+ }
checkpoint-400/trainer_state.json ADDED
@@ -0,0 +1,2419 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.5,
5
+ "eval_steps": 100,
6
+ "global_step": 400,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 5e-06,
14
+ "loss": 1.2856,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 1e-05,
20
+ "loss": 1.2602,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 1.1158,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 2e-05,
32
+ "loss": 1.1952,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 2.5e-05,
38
+ "loss": 1.2183,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.3459,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.04,
49
+ "learning_rate": 3.5e-05,
50
+ "loss": 1.1262,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 4e-05,
56
+ "loss": 0.9534,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 4.5e-05,
62
+ "loss": 1.3216,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 5e-05,
68
+ "loss": 1.1305,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 5.500000000000001e-05,
74
+ "loss": 1.1127,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.07,
79
+ "learning_rate": 6e-05,
80
+ "loss": 0.9814,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 6.500000000000001e-05,
86
+ "loss": 1.2747,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 7e-05,
92
+ "loss": 1.0283,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.09,
97
+ "learning_rate": 7.500000000000001e-05,
98
+ "loss": 1.1814,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 8e-05,
104
+ "loss": 1.0153,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 8.5e-05,
110
+ "loss": 1.075,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.11,
115
+ "learning_rate": 9e-05,
116
+ "loss": 0.9444,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 9.5e-05,
122
+ "loss": 1.0805,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.12,
127
+ "learning_rate": 0.0001,
128
+ "loss": 0.852,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.13,
133
+ "learning_rate": 9.978260869565218e-05,
134
+ "loss": 0.9922,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 0.14,
139
+ "learning_rate": 9.956521739130435e-05,
140
+ "loss": 0.8473,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 0.14,
145
+ "learning_rate": 9.934782608695653e-05,
146
+ "loss": 0.7442,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 0.15,
151
+ "learning_rate": 9.91304347826087e-05,
152
+ "loss": 0.7805,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 0.16,
157
+ "learning_rate": 9.891304347826087e-05,
158
+ "loss": 0.7853,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 0.16,
163
+ "learning_rate": 9.869565217391305e-05,
164
+ "loss": 0.684,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 0.17,
169
+ "learning_rate": 9.847826086956522e-05,
170
+ "loss": 0.6158,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 0.17,
175
+ "learning_rate": 9.82608695652174e-05,
176
+ "loss": 0.6431,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 0.18,
181
+ "learning_rate": 9.804347826086957e-05,
182
+ "loss": 0.6562,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 0.19,
187
+ "learning_rate": 9.782608695652174e-05,
188
+ "loss": 0.5009,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 0.19,
193
+ "learning_rate": 9.760869565217392e-05,
194
+ "loss": 0.3529,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 0.2,
199
+ "learning_rate": 9.739130434782609e-05,
200
+ "loss": 0.409,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 0.21,
205
+ "learning_rate": 9.717391304347827e-05,
206
+ "loss": 0.3635,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 0.21,
211
+ "learning_rate": 9.695652173913044e-05,
212
+ "loss": 0.3853,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 0.22,
217
+ "learning_rate": 9.673913043478261e-05,
218
+ "loss": 0.2333,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 0.23,
223
+ "learning_rate": 9.652173913043479e-05,
224
+ "loss": 0.304,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 0.23,
229
+ "learning_rate": 9.630434782608696e-05,
230
+ "loss": 0.2929,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 0.24,
235
+ "learning_rate": 9.608695652173914e-05,
236
+ "loss": 0.1785,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 0.24,
241
+ "learning_rate": 9.586956521739131e-05,
242
+ "loss": 0.081,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 0.25,
247
+ "learning_rate": 9.565217391304348e-05,
248
+ "loss": 0.4271,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 0.26,
253
+ "learning_rate": 9.543478260869566e-05,
254
+ "loss": 0.1938,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 0.26,
259
+ "learning_rate": 9.521739130434783e-05,
260
+ "loss": 0.2521,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 0.27,
265
+ "learning_rate": 9.5e-05,
266
+ "loss": 0.2784,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 0.28,
271
+ "learning_rate": 9.478260869565218e-05,
272
+ "loss": 0.2935,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 0.28,
277
+ "learning_rate": 9.456521739130435e-05,
278
+ "loss": 0.2826,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 0.29,
283
+ "learning_rate": 9.434782608695653e-05,
284
+ "loss": 0.2279,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 0.29,
289
+ "learning_rate": 9.41304347826087e-05,
290
+ "loss": 0.2364,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 0.3,
295
+ "learning_rate": 9.391304347826087e-05,
296
+ "loss": 0.2543,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 0.31,
301
+ "learning_rate": 9.369565217391305e-05,
302
+ "loss": 0.1922,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 0.31,
307
+ "learning_rate": 9.347826086956522e-05,
308
+ "loss": 0.1936,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 0.32,
313
+ "learning_rate": 9.32608695652174e-05,
314
+ "loss": 0.217,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 0.33,
319
+ "learning_rate": 9.304347826086957e-05,
320
+ "loss": 0.2158,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 0.33,
325
+ "learning_rate": 9.282608695652174e-05,
326
+ "loss": 0.1472,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 0.34,
331
+ "learning_rate": 9.260869565217392e-05,
332
+ "loss": 0.2314,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 0.34,
337
+ "learning_rate": 9.239130434782609e-05,
338
+ "loss": 0.2205,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 0.35,
343
+ "learning_rate": 9.217391304347827e-05,
344
+ "loss": 0.1921,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 0.36,
349
+ "learning_rate": 9.195652173913044e-05,
350
+ "loss": 0.1892,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 0.36,
355
+ "learning_rate": 9.173913043478261e-05,
356
+ "loss": 0.2255,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 0.37,
361
+ "learning_rate": 9.152173913043479e-05,
362
+ "loss": 0.1864,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 0.38,
367
+ "learning_rate": 9.130434782608696e-05,
368
+ "loss": 0.2153,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 0.38,
373
+ "learning_rate": 9.108695652173914e-05,
374
+ "loss": 0.168,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 0.39,
379
+ "learning_rate": 9.086956521739131e-05,
380
+ "loss": 0.0987,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 0.39,
385
+ "learning_rate": 9.065217391304348e-05,
386
+ "loss": 0.2205,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 0.4,
391
+ "learning_rate": 9.043478260869566e-05,
392
+ "loss": 0.119,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 0.41,
397
+ "learning_rate": 9.021739130434783e-05,
398
+ "loss": 0.1378,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 0.41,
403
+ "learning_rate": 9e-05,
404
+ "loss": 0.1345,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 0.42,
409
+ "learning_rate": 8.978260869565218e-05,
410
+ "loss": 0.2253,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 0.42,
415
+ "learning_rate": 8.956521739130435e-05,
416
+ "loss": 0.128,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 0.43,
421
+ "learning_rate": 8.934782608695653e-05,
422
+ "loss": 0.1143,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 0.44,
427
+ "learning_rate": 8.91304347826087e-05,
428
+ "loss": 0.1149,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 0.44,
433
+ "learning_rate": 8.891304347826088e-05,
434
+ "loss": 0.1714,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 0.45,
439
+ "learning_rate": 8.869565217391305e-05,
440
+ "loss": 0.1608,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 0.46,
445
+ "learning_rate": 8.847826086956522e-05,
446
+ "loss": 0.0919,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 0.46,
451
+ "learning_rate": 8.82608695652174e-05,
452
+ "loss": 0.2223,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 0.47,
457
+ "learning_rate": 8.804347826086957e-05,
458
+ "loss": 0.1387,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 0.47,
463
+ "learning_rate": 8.782608695652174e-05,
464
+ "loss": 0.1626,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 0.48,
469
+ "learning_rate": 8.760869565217392e-05,
470
+ "loss": 0.1865,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 0.49,
475
+ "learning_rate": 8.739130434782609e-05,
476
+ "loss": 0.1157,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 0.49,
481
+ "learning_rate": 8.717391304347827e-05,
482
+ "loss": 0.0903,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 0.5,
487
+ "learning_rate": 8.695652173913044e-05,
488
+ "loss": 0.2334,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 0.51,
493
+ "learning_rate": 8.673913043478261e-05,
494
+ "loss": 0.095,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 0.51,
499
+ "learning_rate": 8.652173913043479e-05,
500
+ "loss": 0.1797,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 0.52,
505
+ "learning_rate": 8.630434782608696e-05,
506
+ "loss": 0.1989,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 0.53,
511
+ "learning_rate": 8.608695652173914e-05,
512
+ "loss": 0.1347,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 0.53,
517
+ "learning_rate": 8.586956521739131e-05,
518
+ "loss": 0.1247,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 0.54,
523
+ "learning_rate": 8.565217391304348e-05,
524
+ "loss": 0.2118,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 0.54,
529
+ "learning_rate": 8.543478260869566e-05,
530
+ "loss": 0.1537,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 0.55,
535
+ "learning_rate": 8.521739130434783e-05,
536
+ "loss": 0.3155,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 0.56,
541
+ "learning_rate": 8.5e-05,
542
+ "loss": 0.0565,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 0.56,
547
+ "learning_rate": 8.478260869565218e-05,
548
+ "loss": 0.1259,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 0.57,
553
+ "learning_rate": 8.456521739130435e-05,
554
+ "loss": 0.1834,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 0.57,
559
+ "learning_rate": 8.434782608695653e-05,
560
+ "loss": 0.311,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 0.58,
565
+ "learning_rate": 8.41304347826087e-05,
566
+ "loss": 0.1937,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 0.59,
571
+ "learning_rate": 8.391304347826088e-05,
572
+ "loss": 0.1621,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 0.59,
577
+ "learning_rate": 8.369565217391305e-05,
578
+ "loss": 0.2723,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 0.6,
583
+ "learning_rate": 8.347826086956521e-05,
584
+ "loss": 0.1308,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 0.61,
589
+ "learning_rate": 8.32608695652174e-05,
590
+ "loss": 0.1453,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 0.61,
595
+ "learning_rate": 8.304347826086957e-05,
596
+ "loss": 0.135,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 0.62,
601
+ "learning_rate": 8.282608695652175e-05,
602
+ "loss": 0.1921,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 0.62,
607
+ "learning_rate": 8.260869565217392e-05,
608
+ "loss": 0.0951,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 0.63,
613
+ "learning_rate": 8.23913043478261e-05,
614
+ "loss": 0.1859,
615
+ "step": 101
616
+ },
617
+ {
618
+ "epoch": 0.64,
619
+ "learning_rate": 8.217391304347827e-05,
620
+ "loss": 0.2049,
621
+ "step": 102
622
+ },
623
+ {
624
+ "epoch": 0.64,
625
+ "learning_rate": 8.195652173913044e-05,
626
+ "loss": 0.1328,
627
+ "step": 103
628
+ },
629
+ {
630
+ "epoch": 0.65,
631
+ "learning_rate": 8.173913043478262e-05,
632
+ "loss": 0.1124,
633
+ "step": 104
634
+ },
635
+ {
636
+ "epoch": 0.66,
637
+ "learning_rate": 8.152173913043478e-05,
638
+ "loss": 0.1932,
639
+ "step": 105
640
+ },
641
+ {
642
+ "epoch": 0.66,
643
+ "learning_rate": 8.130434782608696e-05,
644
+ "loss": 0.1878,
645
+ "step": 106
646
+ },
647
+ {
648
+ "epoch": 0.67,
649
+ "learning_rate": 8.108695652173914e-05,
650
+ "loss": 0.1053,
651
+ "step": 107
652
+ },
653
+ {
654
+ "epoch": 0.68,
655
+ "learning_rate": 8.086956521739131e-05,
656
+ "loss": 0.1761,
657
+ "step": 108
658
+ },
659
+ {
660
+ "epoch": 0.68,
661
+ "learning_rate": 8.065217391304348e-05,
662
+ "loss": 0.1566,
663
+ "step": 109
664
+ },
665
+ {
666
+ "epoch": 0.69,
667
+ "learning_rate": 8.043478260869566e-05,
668
+ "loss": 0.1254,
669
+ "step": 110
670
+ },
671
+ {
672
+ "epoch": 0.69,
673
+ "learning_rate": 8.021739130434783e-05,
674
+ "loss": 0.1144,
675
+ "step": 111
676
+ },
677
+ {
678
+ "epoch": 0.7,
679
+ "learning_rate": 8e-05,
680
+ "loss": 0.1335,
681
+ "step": 112
682
+ },
683
+ {
684
+ "epoch": 0.71,
685
+ "learning_rate": 7.978260869565217e-05,
686
+ "loss": 0.1611,
687
+ "step": 113
688
+ },
689
+ {
690
+ "epoch": 0.71,
691
+ "learning_rate": 7.956521739130434e-05,
692
+ "loss": 0.1283,
693
+ "step": 114
694
+ },
695
+ {
696
+ "epoch": 0.72,
697
+ "learning_rate": 7.934782608695653e-05,
698
+ "loss": 0.1697,
699
+ "step": 115
700
+ },
701
+ {
702
+ "epoch": 0.72,
703
+ "learning_rate": 7.91304347826087e-05,
704
+ "loss": 0.1886,
705
+ "step": 116
706
+ },
707
+ {
708
+ "epoch": 0.73,
709
+ "learning_rate": 7.891304347826088e-05,
710
+ "loss": 0.1882,
711
+ "step": 117
712
+ },
713
+ {
714
+ "epoch": 0.74,
715
+ "learning_rate": 7.869565217391305e-05,
716
+ "loss": 0.1962,
717
+ "step": 118
718
+ },
719
+ {
720
+ "epoch": 0.74,
721
+ "learning_rate": 7.847826086956522e-05,
722
+ "loss": 0.104,
723
+ "step": 119
724
+ },
725
+ {
726
+ "epoch": 0.75,
727
+ "learning_rate": 7.82608695652174e-05,
728
+ "loss": 0.1633,
729
+ "step": 120
730
+ },
731
+ {
732
+ "epoch": 0.76,
733
+ "learning_rate": 7.804347826086957e-05,
734
+ "loss": 0.155,
735
+ "step": 121
736
+ },
737
+ {
738
+ "epoch": 0.76,
739
+ "learning_rate": 7.782608695652173e-05,
740
+ "loss": 0.1456,
741
+ "step": 122
742
+ },
743
+ {
744
+ "epoch": 0.77,
745
+ "learning_rate": 7.76086956521739e-05,
746
+ "loss": 0.1612,
747
+ "step": 123
748
+ },
749
+ {
750
+ "epoch": 0.78,
751
+ "learning_rate": 7.73913043478261e-05,
752
+ "loss": 0.0898,
753
+ "step": 124
754
+ },
755
+ {
756
+ "epoch": 0.78,
757
+ "learning_rate": 7.717391304347827e-05,
758
+ "loss": 0.1639,
759
+ "step": 125
760
+ },
761
+ {
762
+ "epoch": 0.79,
763
+ "learning_rate": 7.695652173913044e-05,
764
+ "loss": 0.1077,
765
+ "step": 126
766
+ },
767
+ {
768
+ "epoch": 0.79,
769
+ "learning_rate": 7.673913043478262e-05,
770
+ "loss": 0.1189,
771
+ "step": 127
772
+ },
773
+ {
774
+ "epoch": 0.8,
775
+ "learning_rate": 7.652173913043479e-05,
776
+ "loss": 0.1671,
777
+ "step": 128
778
+ },
779
+ {
780
+ "epoch": 0.81,
781
+ "learning_rate": 7.630434782608696e-05,
782
+ "loss": 0.1334,
783
+ "step": 129
784
+ },
785
+ {
786
+ "epoch": 0.81,
787
+ "learning_rate": 7.608695652173914e-05,
788
+ "loss": 0.0486,
789
+ "step": 130
790
+ },
791
+ {
792
+ "epoch": 0.82,
793
+ "learning_rate": 7.58695652173913e-05,
794
+ "loss": 0.0935,
795
+ "step": 131
796
+ },
797
+ {
798
+ "epoch": 0.82,
799
+ "learning_rate": 7.565217391304347e-05,
800
+ "loss": 0.1172,
801
+ "step": 132
802
+ },
803
+ {
804
+ "epoch": 0.83,
805
+ "learning_rate": 7.543478260869566e-05,
806
+ "loss": 0.1727,
807
+ "step": 133
808
+ },
809
+ {
810
+ "epoch": 0.84,
811
+ "learning_rate": 7.521739130434783e-05,
812
+ "loss": 0.2791,
813
+ "step": 134
814
+ },
815
+ {
816
+ "epoch": 0.84,
817
+ "learning_rate": 7.500000000000001e-05,
818
+ "loss": 0.101,
819
+ "step": 135
820
+ },
821
+ {
822
+ "epoch": 0.85,
823
+ "learning_rate": 7.478260869565218e-05,
824
+ "loss": 0.0511,
825
+ "step": 136
826
+ },
827
+ {
828
+ "epoch": 0.86,
829
+ "learning_rate": 7.456521739130435e-05,
830
+ "loss": 0.1013,
831
+ "step": 137
832
+ },
833
+ {
834
+ "epoch": 0.86,
835
+ "learning_rate": 7.434782608695653e-05,
836
+ "loss": 0.1812,
837
+ "step": 138
838
+ },
839
+ {
840
+ "epoch": 0.87,
841
+ "learning_rate": 7.41304347826087e-05,
842
+ "loss": 0.179,
843
+ "step": 139
844
+ },
845
+ {
846
+ "epoch": 0.88,
847
+ "learning_rate": 7.391304347826086e-05,
848
+ "loss": 0.1382,
849
+ "step": 140
850
+ },
851
+ {
852
+ "epoch": 0.88,
853
+ "learning_rate": 7.369565217391304e-05,
854
+ "loss": 0.0617,
855
+ "step": 141
856
+ },
857
+ {
858
+ "epoch": 0.89,
859
+ "learning_rate": 7.347826086956522e-05,
860
+ "loss": 0.0953,
861
+ "step": 142
862
+ },
863
+ {
864
+ "epoch": 0.89,
865
+ "learning_rate": 7.32608695652174e-05,
866
+ "loss": 0.1096,
867
+ "step": 143
868
+ },
869
+ {
870
+ "epoch": 0.9,
871
+ "learning_rate": 7.304347826086957e-05,
872
+ "loss": 0.0872,
873
+ "step": 144
874
+ },
875
+ {
876
+ "epoch": 0.91,
877
+ "learning_rate": 7.282608695652175e-05,
878
+ "loss": 0.1359,
879
+ "step": 145
880
+ },
881
+ {
882
+ "epoch": 0.91,
883
+ "learning_rate": 7.260869565217392e-05,
884
+ "loss": 0.0824,
885
+ "step": 146
886
+ },
887
+ {
888
+ "epoch": 0.92,
889
+ "learning_rate": 7.23913043478261e-05,
890
+ "loss": 0.1948,
891
+ "step": 147
892
+ },
893
+ {
894
+ "epoch": 0.93,
895
+ "learning_rate": 7.217391304347827e-05,
896
+ "loss": 0.166,
897
+ "step": 148
898
+ },
899
+ {
900
+ "epoch": 0.93,
901
+ "learning_rate": 7.195652173913043e-05,
902
+ "loss": 0.1802,
903
+ "step": 149
904
+ },
905
+ {
906
+ "epoch": 0.94,
907
+ "learning_rate": 7.17391304347826e-05,
908
+ "loss": 0.0748,
909
+ "step": 150
910
+ },
911
+ {
912
+ "epoch": 0.94,
913
+ "learning_rate": 7.152173913043479e-05,
914
+ "loss": 0.1022,
915
+ "step": 151
916
+ },
917
+ {
918
+ "epoch": 0.95,
919
+ "learning_rate": 7.130434782608696e-05,
920
+ "loss": 0.1456,
921
+ "step": 152
922
+ },
923
+ {
924
+ "epoch": 0.96,
925
+ "learning_rate": 7.108695652173914e-05,
926
+ "loss": 0.088,
927
+ "step": 153
928
+ },
929
+ {
930
+ "epoch": 0.96,
931
+ "learning_rate": 7.086956521739131e-05,
932
+ "loss": 0.2109,
933
+ "step": 154
934
+ },
935
+ {
936
+ "epoch": 0.97,
937
+ "learning_rate": 7.065217391304349e-05,
938
+ "loss": 0.0573,
939
+ "step": 155
940
+ },
941
+ {
942
+ "epoch": 0.97,
943
+ "learning_rate": 7.043478260869566e-05,
944
+ "loss": 0.0671,
945
+ "step": 156
946
+ },
947
+ {
948
+ "epoch": 0.98,
949
+ "learning_rate": 7.021739130434783e-05,
950
+ "loss": 0.2151,
951
+ "step": 157
952
+ },
953
+ {
954
+ "epoch": 0.99,
955
+ "learning_rate": 7e-05,
956
+ "loss": 0.1243,
957
+ "step": 158
958
+ },
959
+ {
960
+ "epoch": 0.99,
961
+ "learning_rate": 6.978260869565217e-05,
962
+ "loss": 0.2758,
963
+ "step": 159
964
+ },
965
+ {
966
+ "epoch": 1.0,
967
+ "learning_rate": 6.956521739130436e-05,
968
+ "loss": 0.1371,
969
+ "step": 160
970
+ },
971
+ {
972
+ "epoch": 1.01,
973
+ "learning_rate": 6.934782608695653e-05,
974
+ "loss": 0.1457,
975
+ "step": 161
976
+ },
977
+ {
978
+ "epoch": 1.01,
979
+ "learning_rate": 6.91304347826087e-05,
980
+ "loss": 0.0562,
981
+ "step": 162
982
+ },
983
+ {
984
+ "epoch": 1.02,
985
+ "learning_rate": 6.891304347826088e-05,
986
+ "loss": 0.1361,
987
+ "step": 163
988
+ },
989
+ {
990
+ "epoch": 1.02,
991
+ "learning_rate": 6.869565217391305e-05,
992
+ "loss": 0.0893,
993
+ "step": 164
994
+ },
995
+ {
996
+ "epoch": 1.03,
997
+ "learning_rate": 6.847826086956522e-05,
998
+ "loss": 0.1177,
999
+ "step": 165
1000
+ },
1001
+ {
1002
+ "epoch": 1.04,
1003
+ "learning_rate": 6.826086956521739e-05,
1004
+ "loss": 0.1483,
1005
+ "step": 166
1006
+ },
1007
+ {
1008
+ "epoch": 1.04,
1009
+ "learning_rate": 6.804347826086956e-05,
1010
+ "loss": 0.1176,
1011
+ "step": 167
1012
+ },
1013
+ {
1014
+ "epoch": 1.05,
1015
+ "learning_rate": 6.782608695652173e-05,
1016
+ "loss": 0.162,
1017
+ "step": 168
1018
+ },
1019
+ {
1020
+ "epoch": 1.06,
1021
+ "learning_rate": 6.760869565217392e-05,
1022
+ "loss": 0.0821,
1023
+ "step": 169
1024
+ },
1025
+ {
1026
+ "epoch": 1.06,
1027
+ "learning_rate": 6.73913043478261e-05,
1028
+ "loss": 0.0605,
1029
+ "step": 170
1030
+ },
1031
+ {
1032
+ "epoch": 1.07,
1033
+ "learning_rate": 6.717391304347827e-05,
1034
+ "loss": 0.1529,
1035
+ "step": 171
1036
+ },
1037
+ {
1038
+ "epoch": 1.07,
1039
+ "learning_rate": 6.695652173913044e-05,
1040
+ "loss": 0.1352,
1041
+ "step": 172
1042
+ },
1043
+ {
1044
+ "epoch": 1.08,
1045
+ "learning_rate": 6.673913043478262e-05,
1046
+ "loss": 0.1002,
1047
+ "step": 173
1048
+ },
1049
+ {
1050
+ "epoch": 1.09,
1051
+ "learning_rate": 6.652173913043479e-05,
1052
+ "loss": 0.1243,
1053
+ "step": 174
1054
+ },
1055
+ {
1056
+ "epoch": 1.09,
1057
+ "learning_rate": 6.630434782608695e-05,
1058
+ "loss": 0.1483,
1059
+ "step": 175
1060
+ },
1061
+ {
1062
+ "epoch": 1.1,
1063
+ "learning_rate": 6.608695652173912e-05,
1064
+ "loss": 0.1103,
1065
+ "step": 176
1066
+ },
1067
+ {
1068
+ "epoch": 1.11,
1069
+ "learning_rate": 6.58695652173913e-05,
1070
+ "loss": 0.112,
1071
+ "step": 177
1072
+ },
1073
+ {
1074
+ "epoch": 1.11,
1075
+ "learning_rate": 6.565217391304349e-05,
1076
+ "loss": 0.09,
1077
+ "step": 178
1078
+ },
1079
+ {
1080
+ "epoch": 1.12,
1081
+ "learning_rate": 6.543478260869566e-05,
1082
+ "loss": 0.0639,
1083
+ "step": 179
1084
+ },
1085
+ {
1086
+ "epoch": 1.12,
1087
+ "learning_rate": 6.521739130434783e-05,
1088
+ "loss": 0.0873,
1089
+ "step": 180
1090
+ },
1091
+ {
1092
+ "epoch": 1.13,
1093
+ "learning_rate": 6.500000000000001e-05,
1094
+ "loss": 0.0854,
1095
+ "step": 181
1096
+ },
1097
+ {
1098
+ "epoch": 1.14,
1099
+ "learning_rate": 6.478260869565218e-05,
1100
+ "loss": 0.1007,
1101
+ "step": 182
1102
+ },
1103
+ {
1104
+ "epoch": 1.14,
1105
+ "learning_rate": 6.456521739130436e-05,
1106
+ "loss": 0.1791,
1107
+ "step": 183
1108
+ },
1109
+ {
1110
+ "epoch": 1.15,
1111
+ "learning_rate": 6.434782608695652e-05,
1112
+ "loss": 0.1084,
1113
+ "step": 184
1114
+ },
1115
+ {
1116
+ "epoch": 1.16,
1117
+ "learning_rate": 6.413043478260869e-05,
1118
+ "loss": 0.0995,
1119
+ "step": 185
1120
+ },
1121
+ {
1122
+ "epoch": 1.16,
1123
+ "learning_rate": 6.391304347826086e-05,
1124
+ "loss": 0.0831,
1125
+ "step": 186
1126
+ },
1127
+ {
1128
+ "epoch": 1.17,
1129
+ "learning_rate": 6.369565217391305e-05,
1130
+ "loss": 0.1019,
1131
+ "step": 187
1132
+ },
1133
+ {
1134
+ "epoch": 1.18,
1135
+ "learning_rate": 6.347826086956523e-05,
1136
+ "loss": 0.1882,
1137
+ "step": 188
1138
+ },
1139
+ {
1140
+ "epoch": 1.18,
1141
+ "learning_rate": 6.32608695652174e-05,
1142
+ "loss": 0.1135,
1143
+ "step": 189
1144
+ },
1145
+ {
1146
+ "epoch": 1.19,
1147
+ "learning_rate": 6.304347826086957e-05,
1148
+ "loss": 0.1522,
1149
+ "step": 190
1150
+ },
1151
+ {
1152
+ "epoch": 1.19,
1153
+ "learning_rate": 6.282608695652175e-05,
1154
+ "loss": 0.0972,
1155
+ "step": 191
1156
+ },
1157
+ {
1158
+ "epoch": 1.2,
1159
+ "learning_rate": 6.260869565217392e-05,
1160
+ "loss": 0.0994,
1161
+ "step": 192
1162
+ },
1163
+ {
1164
+ "epoch": 1.21,
1165
+ "learning_rate": 6.239130434782608e-05,
1166
+ "loss": 0.0951,
1167
+ "step": 193
1168
+ },
1169
+ {
1170
+ "epoch": 1.21,
1171
+ "learning_rate": 6.217391304347826e-05,
1172
+ "loss": 0.0671,
1173
+ "step": 194
1174
+ },
1175
+ {
1176
+ "epoch": 1.22,
1177
+ "learning_rate": 6.195652173913043e-05,
1178
+ "loss": 0.0917,
1179
+ "step": 195
1180
+ },
1181
+ {
1182
+ "epoch": 1.23,
1183
+ "learning_rate": 6.173913043478262e-05,
1184
+ "loss": 0.1442,
1185
+ "step": 196
1186
+ },
1187
+ {
1188
+ "epoch": 1.23,
1189
+ "learning_rate": 6.152173913043479e-05,
1190
+ "loss": 0.069,
1191
+ "step": 197
1192
+ },
1193
+ {
1194
+ "epoch": 1.24,
1195
+ "learning_rate": 6.130434782608696e-05,
1196
+ "loss": 0.1055,
1197
+ "step": 198
1198
+ },
1199
+ {
1200
+ "epoch": 1.24,
1201
+ "learning_rate": 6.108695652173914e-05,
1202
+ "loss": 0.1078,
1203
+ "step": 199
1204
+ },
1205
+ {
1206
+ "epoch": 1.25,
1207
+ "learning_rate": 6.086956521739131e-05,
1208
+ "loss": 0.147,
1209
+ "step": 200
1210
+ },
1211
+ {
1212
+ "epoch": 1.26,
1213
+ "learning_rate": 6.0652173913043487e-05,
1214
+ "loss": 0.1841,
1215
+ "step": 201
1216
+ },
1217
+ {
1218
+ "epoch": 1.26,
1219
+ "learning_rate": 6.0434782608695654e-05,
1220
+ "loss": 0.1241,
1221
+ "step": 202
1222
+ },
1223
+ {
1224
+ "epoch": 1.27,
1225
+ "learning_rate": 6.021739130434783e-05,
1226
+ "loss": 0.12,
1227
+ "step": 203
1228
+ },
1229
+ {
1230
+ "epoch": 1.27,
1231
+ "learning_rate": 6e-05,
1232
+ "loss": 0.1256,
1233
+ "step": 204
1234
+ },
1235
+ {
1236
+ "epoch": 1.28,
1237
+ "learning_rate": 5.9782608695652175e-05,
1238
+ "loss": 0.0793,
1239
+ "step": 205
1240
+ },
1241
+ {
1242
+ "epoch": 1.29,
1243
+ "learning_rate": 5.956521739130435e-05,
1244
+ "loss": 0.0887,
1245
+ "step": 206
1246
+ },
1247
+ {
1248
+ "epoch": 1.29,
1249
+ "learning_rate": 5.934782608695652e-05,
1250
+ "loss": 0.1211,
1251
+ "step": 207
1252
+ },
1253
+ {
1254
+ "epoch": 1.3,
1255
+ "learning_rate": 5.9130434782608704e-05,
1256
+ "loss": 0.1238,
1257
+ "step": 208
1258
+ },
1259
+ {
1260
+ "epoch": 1.31,
1261
+ "learning_rate": 5.891304347826088e-05,
1262
+ "loss": 0.1484,
1263
+ "step": 209
1264
+ },
1265
+ {
1266
+ "epoch": 1.31,
1267
+ "learning_rate": 5.869565217391305e-05,
1268
+ "loss": 0.0768,
1269
+ "step": 210
1270
+ },
1271
+ {
1272
+ "epoch": 1.32,
1273
+ "learning_rate": 5.847826086956521e-05,
1274
+ "loss": 0.2152,
1275
+ "step": 211
1276
+ },
1277
+ {
1278
+ "epoch": 1.32,
1279
+ "learning_rate": 5.826086956521739e-05,
1280
+ "loss": 0.1132,
1281
+ "step": 212
1282
+ },
1283
+ {
1284
+ "epoch": 1.33,
1285
+ "learning_rate": 5.804347826086957e-05,
1286
+ "loss": 0.1141,
1287
+ "step": 213
1288
+ },
1289
+ {
1290
+ "epoch": 1.34,
1291
+ "learning_rate": 5.782608695652174e-05,
1292
+ "loss": 0.0904,
1293
+ "step": 214
1294
+ },
1295
+ {
1296
+ "epoch": 1.34,
1297
+ "learning_rate": 5.7608695652173915e-05,
1298
+ "loss": 0.0334,
1299
+ "step": 215
1300
+ },
1301
+ {
1302
+ "epoch": 1.35,
1303
+ "learning_rate": 5.739130434782609e-05,
1304
+ "loss": 0.0939,
1305
+ "step": 216
1306
+ },
1307
+ {
1308
+ "epoch": 1.36,
1309
+ "learning_rate": 5.717391304347827e-05,
1310
+ "loss": 0.1181,
1311
+ "step": 217
1312
+ },
1313
+ {
1314
+ "epoch": 1.36,
1315
+ "learning_rate": 5.695652173913044e-05,
1316
+ "loss": 0.101,
1317
+ "step": 218
1318
+ },
1319
+ {
1320
+ "epoch": 1.37,
1321
+ "learning_rate": 5.673913043478262e-05,
1322
+ "loss": 0.1756,
1323
+ "step": 219
1324
+ },
1325
+ {
1326
+ "epoch": 1.38,
1327
+ "learning_rate": 5.652173913043478e-05,
1328
+ "loss": 0.1129,
1329
+ "step": 220
1330
+ },
1331
+ {
1332
+ "epoch": 1.38,
1333
+ "learning_rate": 5.630434782608696e-05,
1334
+ "loss": 0.0689,
1335
+ "step": 221
1336
+ },
1337
+ {
1338
+ "epoch": 1.39,
1339
+ "learning_rate": 5.608695652173913e-05,
1340
+ "loss": 0.094,
1341
+ "step": 222
1342
+ },
1343
+ {
1344
+ "epoch": 1.39,
1345
+ "learning_rate": 5.5869565217391306e-05,
1346
+ "loss": 0.1357,
1347
+ "step": 223
1348
+ },
1349
+ {
1350
+ "epoch": 1.4,
1351
+ "learning_rate": 5.565217391304348e-05,
1352
+ "loss": 0.0742,
1353
+ "step": 224
1354
+ },
1355
+ {
1356
+ "epoch": 1.41,
1357
+ "learning_rate": 5.5434782608695654e-05,
1358
+ "loss": 0.0943,
1359
+ "step": 225
1360
+ },
1361
+ {
1362
+ "epoch": 1.41,
1363
+ "learning_rate": 5.5217391304347835e-05,
1364
+ "loss": 0.1144,
1365
+ "step": 226
1366
+ },
1367
+ {
1368
+ "epoch": 1.42,
1369
+ "learning_rate": 5.500000000000001e-05,
1370
+ "loss": 0.1045,
1371
+ "step": 227
1372
+ },
1373
+ {
1374
+ "epoch": 1.43,
1375
+ "learning_rate": 5.478260869565217e-05,
1376
+ "loss": 0.129,
1377
+ "step": 228
1378
+ },
1379
+ {
1380
+ "epoch": 1.43,
1381
+ "learning_rate": 5.456521739130434e-05,
1382
+ "loss": 0.1173,
1383
+ "step": 229
1384
+ },
1385
+ {
1386
+ "epoch": 1.44,
1387
+ "learning_rate": 5.4347826086956524e-05,
1388
+ "loss": 0.1838,
1389
+ "step": 230
1390
+ },
1391
+ {
1392
+ "epoch": 1.44,
1393
+ "learning_rate": 5.41304347826087e-05,
1394
+ "loss": 0.1537,
1395
+ "step": 231
1396
+ },
1397
+ {
1398
+ "epoch": 1.45,
1399
+ "learning_rate": 5.391304347826087e-05,
1400
+ "loss": 0.078,
1401
+ "step": 232
1402
+ },
1403
+ {
1404
+ "epoch": 1.46,
1405
+ "learning_rate": 5.3695652173913046e-05,
1406
+ "loss": 0.1471,
1407
+ "step": 233
1408
+ },
1409
+ {
1410
+ "epoch": 1.46,
1411
+ "learning_rate": 5.347826086956522e-05,
1412
+ "loss": 0.1094,
1413
+ "step": 234
1414
+ },
1415
+ {
1416
+ "epoch": 1.47,
1417
+ "learning_rate": 5.32608695652174e-05,
1418
+ "loss": 0.1005,
1419
+ "step": 235
1420
+ },
1421
+ {
1422
+ "epoch": 1.48,
1423
+ "learning_rate": 5.3043478260869574e-05,
1424
+ "loss": 0.1354,
1425
+ "step": 236
1426
+ },
1427
+ {
1428
+ "epoch": 1.48,
1429
+ "learning_rate": 5.2826086956521735e-05,
1430
+ "loss": 0.0875,
1431
+ "step": 237
1432
+ },
1433
+ {
1434
+ "epoch": 1.49,
1435
+ "learning_rate": 5.260869565217391e-05,
1436
+ "loss": 0.1438,
1437
+ "step": 238
1438
+ },
1439
+ {
1440
+ "epoch": 1.49,
1441
+ "learning_rate": 5.239130434782609e-05,
1442
+ "loss": 0.1666,
1443
+ "step": 239
1444
+ },
1445
+ {
1446
+ "epoch": 1.5,
1447
+ "learning_rate": 5.217391304347826e-05,
1448
+ "loss": 0.0433,
1449
+ "step": 240
1450
+ },
1451
+ {
1452
+ "epoch": 1.51,
1453
+ "learning_rate": 5.195652173913044e-05,
1454
+ "loss": 0.0899,
1455
+ "step": 241
1456
+ },
1457
+ {
1458
+ "epoch": 1.51,
1459
+ "learning_rate": 5.173913043478261e-05,
1460
+ "loss": 0.1092,
1461
+ "step": 242
1462
+ },
1463
+ {
1464
+ "epoch": 1.52,
1465
+ "learning_rate": 5.1521739130434785e-05,
1466
+ "loss": 0.0806,
1467
+ "step": 243
1468
+ },
1469
+ {
1470
+ "epoch": 1.52,
1471
+ "learning_rate": 5.1304347826086966e-05,
1472
+ "loss": 0.1033,
1473
+ "step": 244
1474
+ },
1475
+ {
1476
+ "epoch": 1.53,
1477
+ "learning_rate": 5.108695652173914e-05,
1478
+ "loss": 0.107,
1479
+ "step": 245
1480
+ },
1481
+ {
1482
+ "epoch": 1.54,
1483
+ "learning_rate": 5.08695652173913e-05,
1484
+ "loss": 0.1478,
1485
+ "step": 246
1486
+ },
1487
+ {
1488
+ "epoch": 1.54,
1489
+ "learning_rate": 5.0652173913043474e-05,
1490
+ "loss": 0.1734,
1491
+ "step": 247
1492
+ },
1493
+ {
1494
+ "epoch": 1.55,
1495
+ "learning_rate": 5.0434782608695655e-05,
1496
+ "loss": 0.065,
1497
+ "step": 248
1498
+ },
1499
+ {
1500
+ "epoch": 1.56,
1501
+ "learning_rate": 5.021739130434783e-05,
1502
+ "loss": 0.1782,
1503
+ "step": 249
1504
+ },
1505
+ {
1506
+ "epoch": 1.56,
1507
+ "learning_rate": 5e-05,
1508
+ "loss": 0.1212,
1509
+ "step": 250
1510
+ },
1511
+ {
1512
+ "epoch": 1.57,
1513
+ "learning_rate": 4.9782608695652176e-05,
1514
+ "loss": 0.1234,
1515
+ "step": 251
1516
+ },
1517
+ {
1518
+ "epoch": 1.57,
1519
+ "learning_rate": 4.956521739130435e-05,
1520
+ "loss": 0.0608,
1521
+ "step": 252
1522
+ },
1523
+ {
1524
+ "epoch": 1.58,
1525
+ "learning_rate": 4.9347826086956524e-05,
1526
+ "loss": 0.0719,
1527
+ "step": 253
1528
+ },
1529
+ {
1530
+ "epoch": 1.59,
1531
+ "learning_rate": 4.91304347826087e-05,
1532
+ "loss": 0.1053,
1533
+ "step": 254
1534
+ },
1535
+ {
1536
+ "epoch": 1.59,
1537
+ "learning_rate": 4.891304347826087e-05,
1538
+ "loss": 0.0209,
1539
+ "step": 255
1540
+ },
1541
+ {
1542
+ "epoch": 1.6,
1543
+ "learning_rate": 4.8695652173913046e-05,
1544
+ "loss": 0.1123,
1545
+ "step": 256
1546
+ },
1547
+ {
1548
+ "epoch": 1.61,
1549
+ "learning_rate": 4.847826086956522e-05,
1550
+ "loss": 0.0561,
1551
+ "step": 257
1552
+ },
1553
+ {
1554
+ "epoch": 1.61,
1555
+ "learning_rate": 4.8260869565217394e-05,
1556
+ "loss": 0.0704,
1557
+ "step": 258
1558
+ },
1559
+ {
1560
+ "epoch": 1.62,
1561
+ "learning_rate": 4.804347826086957e-05,
1562
+ "loss": 0.1738,
1563
+ "step": 259
1564
+ },
1565
+ {
1566
+ "epoch": 1.62,
1567
+ "learning_rate": 4.782608695652174e-05,
1568
+ "loss": 0.1209,
1569
+ "step": 260
1570
+ },
1571
+ {
1572
+ "epoch": 1.63,
1573
+ "learning_rate": 4.7608695652173916e-05,
1574
+ "loss": 0.1059,
1575
+ "step": 261
1576
+ },
1577
+ {
1578
+ "epoch": 1.64,
1579
+ "learning_rate": 4.739130434782609e-05,
1580
+ "loss": 0.0983,
1581
+ "step": 262
1582
+ },
1583
+ {
1584
+ "epoch": 1.64,
1585
+ "learning_rate": 4.7173913043478264e-05,
1586
+ "loss": 0.0919,
1587
+ "step": 263
1588
+ },
1589
+ {
1590
+ "epoch": 1.65,
1591
+ "learning_rate": 4.695652173913044e-05,
1592
+ "loss": 0.06,
1593
+ "step": 264
1594
+ },
1595
+ {
1596
+ "epoch": 1.66,
1597
+ "learning_rate": 4.673913043478261e-05,
1598
+ "loss": 0.0488,
1599
+ "step": 265
1600
+ },
1601
+ {
1602
+ "epoch": 1.66,
1603
+ "learning_rate": 4.6521739130434785e-05,
1604
+ "loss": 0.1182,
1605
+ "step": 266
1606
+ },
1607
+ {
1608
+ "epoch": 1.67,
1609
+ "learning_rate": 4.630434782608696e-05,
1610
+ "loss": 0.0696,
1611
+ "step": 267
1612
+ },
1613
+ {
1614
+ "epoch": 1.68,
1615
+ "learning_rate": 4.608695652173913e-05,
1616
+ "loss": 0.1323,
1617
+ "step": 268
1618
+ },
1619
+ {
1620
+ "epoch": 1.68,
1621
+ "learning_rate": 4.586956521739131e-05,
1622
+ "loss": 0.1429,
1623
+ "step": 269
1624
+ },
1625
+ {
1626
+ "epoch": 1.69,
1627
+ "learning_rate": 4.565217391304348e-05,
1628
+ "loss": 0.0536,
1629
+ "step": 270
1630
+ },
1631
+ {
1632
+ "epoch": 1.69,
1633
+ "learning_rate": 4.5434782608695655e-05,
1634
+ "loss": 0.1794,
1635
+ "step": 271
1636
+ },
1637
+ {
1638
+ "epoch": 1.7,
1639
+ "learning_rate": 4.521739130434783e-05,
1640
+ "loss": 0.1623,
1641
+ "step": 272
1642
+ },
1643
+ {
1644
+ "epoch": 1.71,
1645
+ "learning_rate": 4.5e-05,
1646
+ "loss": 0.1291,
1647
+ "step": 273
1648
+ },
1649
+ {
1650
+ "epoch": 1.71,
1651
+ "learning_rate": 4.478260869565218e-05,
1652
+ "loss": 0.0282,
1653
+ "step": 274
1654
+ },
1655
+ {
1656
+ "epoch": 1.72,
1657
+ "learning_rate": 4.456521739130435e-05,
1658
+ "loss": 0.121,
1659
+ "step": 275
1660
+ },
1661
+ {
1662
+ "epoch": 1.73,
1663
+ "learning_rate": 4.4347826086956525e-05,
1664
+ "loss": 0.0357,
1665
+ "step": 276
1666
+ },
1667
+ {
1668
+ "epoch": 1.73,
1669
+ "learning_rate": 4.41304347826087e-05,
1670
+ "loss": 0.2515,
1671
+ "step": 277
1672
+ },
1673
+ {
1674
+ "epoch": 1.74,
1675
+ "learning_rate": 4.391304347826087e-05,
1676
+ "loss": 0.0873,
1677
+ "step": 278
1678
+ },
1679
+ {
1680
+ "epoch": 1.74,
1681
+ "learning_rate": 4.3695652173913046e-05,
1682
+ "loss": 0.1023,
1683
+ "step": 279
1684
+ },
1685
+ {
1686
+ "epoch": 1.75,
1687
+ "learning_rate": 4.347826086956522e-05,
1688
+ "loss": 0.0918,
1689
+ "step": 280
1690
+ },
1691
+ {
1692
+ "epoch": 1.76,
1693
+ "learning_rate": 4.3260869565217394e-05,
1694
+ "loss": 0.1123,
1695
+ "step": 281
1696
+ },
1697
+ {
1698
+ "epoch": 1.76,
1699
+ "learning_rate": 4.304347826086957e-05,
1700
+ "loss": 0.1115,
1701
+ "step": 282
1702
+ },
1703
+ {
1704
+ "epoch": 1.77,
1705
+ "learning_rate": 4.282608695652174e-05,
1706
+ "loss": 0.0779,
1707
+ "step": 283
1708
+ },
1709
+ {
1710
+ "epoch": 1.77,
1711
+ "learning_rate": 4.2608695652173916e-05,
1712
+ "loss": 0.1158,
1713
+ "step": 284
1714
+ },
1715
+ {
1716
+ "epoch": 1.78,
1717
+ "learning_rate": 4.239130434782609e-05,
1718
+ "loss": 0.1626,
1719
+ "step": 285
1720
+ },
1721
+ {
1722
+ "epoch": 1.79,
1723
+ "learning_rate": 4.2173913043478264e-05,
1724
+ "loss": 0.0482,
1725
+ "step": 286
1726
+ },
1727
+ {
1728
+ "epoch": 1.79,
1729
+ "learning_rate": 4.195652173913044e-05,
1730
+ "loss": 0.0639,
1731
+ "step": 287
1732
+ },
1733
+ {
1734
+ "epoch": 1.8,
1735
+ "learning_rate": 4.1739130434782605e-05,
1736
+ "loss": 0.13,
1737
+ "step": 288
1738
+ },
1739
+ {
1740
+ "epoch": 1.81,
1741
+ "learning_rate": 4.1521739130434786e-05,
1742
+ "loss": 0.0602,
1743
+ "step": 289
1744
+ },
1745
+ {
1746
+ "epoch": 1.81,
1747
+ "learning_rate": 4.130434782608696e-05,
1748
+ "loss": 0.1464,
1749
+ "step": 290
1750
+ },
1751
+ {
1752
+ "epoch": 1.82,
1753
+ "learning_rate": 4.1086956521739134e-05,
1754
+ "loss": 0.0937,
1755
+ "step": 291
1756
+ },
1757
+ {
1758
+ "epoch": 1.82,
1759
+ "learning_rate": 4.086956521739131e-05,
1760
+ "loss": 0.0836,
1761
+ "step": 292
1762
+ },
1763
+ {
1764
+ "epoch": 1.83,
1765
+ "learning_rate": 4.065217391304348e-05,
1766
+ "loss": 0.0842,
1767
+ "step": 293
1768
+ },
1769
+ {
1770
+ "epoch": 1.84,
1771
+ "learning_rate": 4.0434782608695655e-05,
1772
+ "loss": 0.1501,
1773
+ "step": 294
1774
+ },
1775
+ {
1776
+ "epoch": 1.84,
1777
+ "learning_rate": 4.021739130434783e-05,
1778
+ "loss": 0.0968,
1779
+ "step": 295
1780
+ },
1781
+ {
1782
+ "epoch": 1.85,
1783
+ "learning_rate": 4e-05,
1784
+ "loss": 0.1041,
1785
+ "step": 296
1786
+ },
1787
+ {
1788
+ "epoch": 1.86,
1789
+ "learning_rate": 3.978260869565217e-05,
1790
+ "loss": 0.096,
1791
+ "step": 297
1792
+ },
1793
+ {
1794
+ "epoch": 1.86,
1795
+ "learning_rate": 3.956521739130435e-05,
1796
+ "loss": 0.0934,
1797
+ "step": 298
1798
+ },
1799
+ {
1800
+ "epoch": 1.87,
1801
+ "learning_rate": 3.9347826086956525e-05,
1802
+ "loss": 0.0797,
1803
+ "step": 299
1804
+ },
1805
+ {
1806
+ "epoch": 1.88,
1807
+ "learning_rate": 3.91304347826087e-05,
1808
+ "loss": 0.098,
1809
+ "step": 300
1810
+ },
1811
+ {
1812
+ "epoch": 1.88,
1813
+ "learning_rate": 3.8913043478260866e-05,
1814
+ "loss": 0.0695,
1815
+ "step": 301
1816
+ },
1817
+ {
1818
+ "epoch": 1.89,
1819
+ "learning_rate": 3.869565217391305e-05,
1820
+ "loss": 0.1073,
1821
+ "step": 302
1822
+ },
1823
+ {
1824
+ "epoch": 1.89,
1825
+ "learning_rate": 3.847826086956522e-05,
1826
+ "loss": 0.1902,
1827
+ "step": 303
1828
+ },
1829
+ {
1830
+ "epoch": 1.9,
1831
+ "learning_rate": 3.8260869565217395e-05,
1832
+ "loss": 0.0692,
1833
+ "step": 304
1834
+ },
1835
+ {
1836
+ "epoch": 1.91,
1837
+ "learning_rate": 3.804347826086957e-05,
1838
+ "loss": 0.0914,
1839
+ "step": 305
1840
+ },
1841
+ {
1842
+ "epoch": 1.91,
1843
+ "learning_rate": 3.7826086956521736e-05,
1844
+ "loss": 0.0945,
1845
+ "step": 306
1846
+ },
1847
+ {
1848
+ "epoch": 1.92,
1849
+ "learning_rate": 3.7608695652173917e-05,
1850
+ "loss": 0.3069,
1851
+ "step": 307
1852
+ },
1853
+ {
1854
+ "epoch": 1.93,
1855
+ "learning_rate": 3.739130434782609e-05,
1856
+ "loss": 0.0858,
1857
+ "step": 308
1858
+ },
1859
+ {
1860
+ "epoch": 1.93,
1861
+ "learning_rate": 3.7173913043478264e-05,
1862
+ "loss": 0.0798,
1863
+ "step": 309
1864
+ },
1865
+ {
1866
+ "epoch": 1.94,
1867
+ "learning_rate": 3.695652173913043e-05,
1868
+ "loss": 0.0423,
1869
+ "step": 310
1870
+ },
1871
+ {
1872
+ "epoch": 1.94,
1873
+ "learning_rate": 3.673913043478261e-05,
1874
+ "loss": 0.0677,
1875
+ "step": 311
1876
+ },
1877
+ {
1878
+ "epoch": 1.95,
1879
+ "learning_rate": 3.6521739130434786e-05,
1880
+ "loss": 0.0816,
1881
+ "step": 312
1882
+ },
1883
+ {
1884
+ "epoch": 1.96,
1885
+ "learning_rate": 3.630434782608696e-05,
1886
+ "loss": 0.0573,
1887
+ "step": 313
1888
+ },
1889
+ {
1890
+ "epoch": 1.96,
1891
+ "learning_rate": 3.6086956521739134e-05,
1892
+ "loss": 0.059,
1893
+ "step": 314
1894
+ },
1895
+ {
1896
+ "epoch": 1.97,
1897
+ "learning_rate": 3.58695652173913e-05,
1898
+ "loss": 0.0966,
1899
+ "step": 315
1900
+ },
1901
+ {
1902
+ "epoch": 1.98,
1903
+ "learning_rate": 3.565217391304348e-05,
1904
+ "loss": 0.1021,
1905
+ "step": 316
1906
+ },
1907
+ {
1908
+ "epoch": 1.98,
1909
+ "learning_rate": 3.5434782608695656e-05,
1910
+ "loss": 0.0716,
1911
+ "step": 317
1912
+ },
1913
+ {
1914
+ "epoch": 1.99,
1915
+ "learning_rate": 3.521739130434783e-05,
1916
+ "loss": 0.1406,
1917
+ "step": 318
1918
+ },
1919
+ {
1920
+ "epoch": 1.99,
1921
+ "learning_rate": 3.5e-05,
1922
+ "loss": 0.1697,
1923
+ "step": 319
1924
+ },
1925
+ {
1926
+ "epoch": 2.0,
1927
+ "learning_rate": 3.478260869565218e-05,
1928
+ "loss": 0.074,
1929
+ "step": 320
1930
+ },
1931
+ {
1932
+ "epoch": 2.01,
1933
+ "learning_rate": 3.456521739130435e-05,
1934
+ "loss": 0.0922,
1935
+ "step": 321
1936
+ },
1937
+ {
1938
+ "epoch": 2.01,
1939
+ "learning_rate": 3.4347826086956526e-05,
1940
+ "loss": 0.0864,
1941
+ "step": 322
1942
+ },
1943
+ {
1944
+ "epoch": 2.02,
1945
+ "learning_rate": 3.413043478260869e-05,
1946
+ "loss": 0.186,
1947
+ "step": 323
1948
+ },
1949
+ {
1950
+ "epoch": 2.02,
1951
+ "learning_rate": 3.3913043478260867e-05,
1952
+ "loss": 0.0498,
1953
+ "step": 324
1954
+ },
1955
+ {
1956
+ "epoch": 2.03,
1957
+ "learning_rate": 3.369565217391305e-05,
1958
+ "loss": 0.0901,
1959
+ "step": 325
1960
+ },
1961
+ {
1962
+ "epoch": 2.04,
1963
+ "learning_rate": 3.347826086956522e-05,
1964
+ "loss": 0.0907,
1965
+ "step": 326
1966
+ },
1967
+ {
1968
+ "epoch": 2.04,
1969
+ "learning_rate": 3.3260869565217395e-05,
1970
+ "loss": 0.0481,
1971
+ "step": 327
1972
+ },
1973
+ {
1974
+ "epoch": 2.05,
1975
+ "learning_rate": 3.304347826086956e-05,
1976
+ "loss": 0.0889,
1977
+ "step": 328
1978
+ },
1979
+ {
1980
+ "epoch": 2.06,
1981
+ "learning_rate": 3.282608695652174e-05,
1982
+ "loss": 0.1153,
1983
+ "step": 329
1984
+ },
1985
+ {
1986
+ "epoch": 2.06,
1987
+ "learning_rate": 3.260869565217392e-05,
1988
+ "loss": 0.0999,
1989
+ "step": 330
1990
+ },
1991
+ {
1992
+ "epoch": 2.07,
1993
+ "learning_rate": 3.239130434782609e-05,
1994
+ "loss": 0.1522,
1995
+ "step": 331
1996
+ },
1997
+ {
1998
+ "epoch": 2.08,
1999
+ "learning_rate": 3.217391304347826e-05,
2000
+ "loss": 0.0708,
2001
+ "step": 332
2002
+ },
2003
+ {
2004
+ "epoch": 2.08,
2005
+ "learning_rate": 3.195652173913043e-05,
2006
+ "loss": 0.0612,
2007
+ "step": 333
2008
+ },
2009
+ {
2010
+ "epoch": 2.09,
2011
+ "learning_rate": 3.173913043478261e-05,
2012
+ "loss": 0.0888,
2013
+ "step": 334
2014
+ },
2015
+ {
2016
+ "epoch": 2.09,
2017
+ "learning_rate": 3.152173913043479e-05,
2018
+ "loss": 0.1127,
2019
+ "step": 335
2020
+ },
2021
+ {
2022
+ "epoch": 2.1,
2023
+ "learning_rate": 3.130434782608696e-05,
2024
+ "loss": 0.0957,
2025
+ "step": 336
2026
+ },
2027
+ {
2028
+ "epoch": 2.11,
2029
+ "learning_rate": 3.108695652173913e-05,
2030
+ "loss": 0.0678,
2031
+ "step": 337
2032
+ },
2033
+ {
2034
+ "epoch": 2.11,
2035
+ "learning_rate": 3.086956521739131e-05,
2036
+ "loss": 0.0593,
2037
+ "step": 338
2038
+ },
2039
+ {
2040
+ "epoch": 2.12,
2041
+ "learning_rate": 3.065217391304348e-05,
2042
+ "loss": 0.0426,
2043
+ "step": 339
2044
+ },
2045
+ {
2046
+ "epoch": 2.12,
2047
+ "learning_rate": 3.0434782608695656e-05,
2048
+ "loss": 0.0921,
2049
+ "step": 340
2050
+ },
2051
+ {
2052
+ "epoch": 2.13,
2053
+ "learning_rate": 3.0217391304347827e-05,
2054
+ "loss": 0.0893,
2055
+ "step": 341
2056
+ },
2057
+ {
2058
+ "epoch": 2.14,
2059
+ "learning_rate": 3e-05,
2060
+ "loss": 0.0812,
2061
+ "step": 342
2062
+ },
2063
+ {
2064
+ "epoch": 2.14,
2065
+ "learning_rate": 2.9782608695652175e-05,
2066
+ "loss": 0.0792,
2067
+ "step": 343
2068
+ },
2069
+ {
2070
+ "epoch": 2.15,
2071
+ "learning_rate": 2.9565217391304352e-05,
2072
+ "loss": 0.0879,
2073
+ "step": 344
2074
+ },
2075
+ {
2076
+ "epoch": 2.16,
2077
+ "learning_rate": 2.9347826086956526e-05,
2078
+ "loss": 0.1001,
2079
+ "step": 345
2080
+ },
2081
+ {
2082
+ "epoch": 2.16,
2083
+ "learning_rate": 2.9130434782608696e-05,
2084
+ "loss": 0.0404,
2085
+ "step": 346
2086
+ },
2087
+ {
2088
+ "epoch": 2.17,
2089
+ "learning_rate": 2.891304347826087e-05,
2090
+ "loss": 0.085,
2091
+ "step": 347
2092
+ },
2093
+ {
2094
+ "epoch": 2.17,
2095
+ "learning_rate": 2.8695652173913044e-05,
2096
+ "loss": 0.0907,
2097
+ "step": 348
2098
+ },
2099
+ {
2100
+ "epoch": 2.18,
2101
+ "learning_rate": 2.847826086956522e-05,
2102
+ "loss": 0.0475,
2103
+ "step": 349
2104
+ },
2105
+ {
2106
+ "epoch": 2.19,
2107
+ "learning_rate": 2.826086956521739e-05,
2108
+ "loss": 0.0555,
2109
+ "step": 350
2110
+ },
2111
+ {
2112
+ "epoch": 2.19,
2113
+ "learning_rate": 2.8043478260869566e-05,
2114
+ "loss": 0.058,
2115
+ "step": 351
2116
+ },
2117
+ {
2118
+ "epoch": 2.2,
2119
+ "learning_rate": 2.782608695652174e-05,
2120
+ "loss": 0.0339,
2121
+ "step": 352
2122
+ },
2123
+ {
2124
+ "epoch": 2.21,
2125
+ "learning_rate": 2.7608695652173917e-05,
2126
+ "loss": 0.0455,
2127
+ "step": 353
2128
+ },
2129
+ {
2130
+ "epoch": 2.21,
2131
+ "learning_rate": 2.7391304347826085e-05,
2132
+ "loss": 0.1721,
2133
+ "step": 354
2134
+ },
2135
+ {
2136
+ "epoch": 2.22,
2137
+ "learning_rate": 2.7173913043478262e-05,
2138
+ "loss": 0.1153,
2139
+ "step": 355
2140
+ },
2141
+ {
2142
+ "epoch": 2.23,
2143
+ "learning_rate": 2.6956521739130436e-05,
2144
+ "loss": 0.0864,
2145
+ "step": 356
2146
+ },
2147
+ {
2148
+ "epoch": 2.23,
2149
+ "learning_rate": 2.673913043478261e-05,
2150
+ "loss": 0.052,
2151
+ "step": 357
2152
+ },
2153
+ {
2154
+ "epoch": 2.24,
2155
+ "learning_rate": 2.6521739130434787e-05,
2156
+ "loss": 0.1067,
2157
+ "step": 358
2158
+ },
2159
+ {
2160
+ "epoch": 2.24,
2161
+ "learning_rate": 2.6304347826086954e-05,
2162
+ "loss": 0.1457,
2163
+ "step": 359
2164
+ },
2165
+ {
2166
+ "epoch": 2.25,
2167
+ "learning_rate": 2.608695652173913e-05,
2168
+ "loss": 0.0143,
2169
+ "step": 360
2170
+ },
2171
+ {
2172
+ "epoch": 2.26,
2173
+ "learning_rate": 2.5869565217391305e-05,
2174
+ "loss": 0.0317,
2175
+ "step": 361
2176
+ },
2177
+ {
2178
+ "epoch": 2.26,
2179
+ "learning_rate": 2.5652173913043483e-05,
2180
+ "loss": 0.1009,
2181
+ "step": 362
2182
+ },
2183
+ {
2184
+ "epoch": 2.27,
2185
+ "learning_rate": 2.543478260869565e-05,
2186
+ "loss": 0.0647,
2187
+ "step": 363
2188
+ },
2189
+ {
2190
+ "epoch": 2.27,
2191
+ "learning_rate": 2.5217391304347827e-05,
2192
+ "loss": 0.0487,
2193
+ "step": 364
2194
+ },
2195
+ {
2196
+ "epoch": 2.28,
2197
+ "learning_rate": 2.5e-05,
2198
+ "loss": 0.027,
2199
+ "step": 365
2200
+ },
2201
+ {
2202
+ "epoch": 2.29,
2203
+ "learning_rate": 2.4782608695652175e-05,
2204
+ "loss": 0.0654,
2205
+ "step": 366
2206
+ },
2207
+ {
2208
+ "epoch": 2.29,
2209
+ "learning_rate": 2.456521739130435e-05,
2210
+ "loss": 0.0511,
2211
+ "step": 367
2212
+ },
2213
+ {
2214
+ "epoch": 2.3,
2215
+ "learning_rate": 2.4347826086956523e-05,
2216
+ "loss": 0.0678,
2217
+ "step": 368
2218
+ },
2219
+ {
2220
+ "epoch": 2.31,
2221
+ "learning_rate": 2.4130434782608697e-05,
2222
+ "loss": 0.058,
2223
+ "step": 369
2224
+ },
2225
+ {
2226
+ "epoch": 2.31,
2227
+ "learning_rate": 2.391304347826087e-05,
2228
+ "loss": 0.0787,
2229
+ "step": 370
2230
+ },
2231
+ {
2232
+ "epoch": 2.32,
2233
+ "learning_rate": 2.3695652173913045e-05,
2234
+ "loss": 0.056,
2235
+ "step": 371
2236
+ },
2237
+ {
2238
+ "epoch": 2.33,
2239
+ "learning_rate": 2.347826086956522e-05,
2240
+ "loss": 0.1519,
2241
+ "step": 372
2242
+ },
2243
+ {
2244
+ "epoch": 2.33,
2245
+ "learning_rate": 2.3260869565217393e-05,
2246
+ "loss": 0.0411,
2247
+ "step": 373
2248
+ },
2249
+ {
2250
+ "epoch": 2.34,
2251
+ "learning_rate": 2.3043478260869567e-05,
2252
+ "loss": 0.0525,
2253
+ "step": 374
2254
+ },
2255
+ {
2256
+ "epoch": 2.34,
2257
+ "learning_rate": 2.282608695652174e-05,
2258
+ "loss": 0.0942,
2259
+ "step": 375
2260
+ },
2261
+ {
2262
+ "epoch": 2.35,
2263
+ "learning_rate": 2.2608695652173914e-05,
2264
+ "loss": 0.0333,
2265
+ "step": 376
2266
+ },
2267
+ {
2268
+ "epoch": 2.36,
2269
+ "learning_rate": 2.239130434782609e-05,
2270
+ "loss": 0.0497,
2271
+ "step": 377
2272
+ },
2273
+ {
2274
+ "epoch": 2.36,
2275
+ "learning_rate": 2.2173913043478262e-05,
2276
+ "loss": 0.0733,
2277
+ "step": 378
2278
+ },
2279
+ {
2280
+ "epoch": 2.37,
2281
+ "learning_rate": 2.1956521739130436e-05,
2282
+ "loss": 0.1511,
2283
+ "step": 379
2284
+ },
2285
+ {
2286
+ "epoch": 2.38,
2287
+ "learning_rate": 2.173913043478261e-05,
2288
+ "loss": 0.1551,
2289
+ "step": 380
2290
+ },
2291
+ {
2292
+ "epoch": 2.38,
2293
+ "learning_rate": 2.1521739130434784e-05,
2294
+ "loss": 0.0532,
2295
+ "step": 381
2296
+ },
2297
+ {
2298
+ "epoch": 2.39,
2299
+ "learning_rate": 2.1304347826086958e-05,
2300
+ "loss": 0.0677,
2301
+ "step": 382
2302
+ },
2303
+ {
2304
+ "epoch": 2.39,
2305
+ "learning_rate": 2.1086956521739132e-05,
2306
+ "loss": 0.042,
2307
+ "step": 383
2308
+ },
2309
+ {
2310
+ "epoch": 2.4,
2311
+ "learning_rate": 2.0869565217391303e-05,
2312
+ "loss": 0.0878,
2313
+ "step": 384
2314
+ },
2315
+ {
2316
+ "epoch": 2.41,
2317
+ "learning_rate": 2.065217391304348e-05,
2318
+ "loss": 0.1585,
2319
+ "step": 385
2320
+ },
2321
+ {
2322
+ "epoch": 2.41,
2323
+ "learning_rate": 2.0434782608695654e-05,
2324
+ "loss": 0.0931,
2325
+ "step": 386
2326
+ },
2327
+ {
2328
+ "epoch": 2.42,
2329
+ "learning_rate": 2.0217391304347828e-05,
2330
+ "loss": 0.1007,
2331
+ "step": 387
2332
+ },
2333
+ {
2334
+ "epoch": 2.42,
2335
+ "learning_rate": 2e-05,
2336
+ "loss": 0.0597,
2337
+ "step": 388
2338
+ },
2339
+ {
2340
+ "epoch": 2.43,
2341
+ "learning_rate": 1.9782608695652176e-05,
2342
+ "loss": 0.0789,
2343
+ "step": 389
2344
+ },
2345
+ {
2346
+ "epoch": 2.44,
2347
+ "learning_rate": 1.956521739130435e-05,
2348
+ "loss": 0.0721,
2349
+ "step": 390
2350
+ },
2351
+ {
2352
+ "epoch": 2.44,
2353
+ "learning_rate": 1.9347826086956523e-05,
2354
+ "loss": 0.1328,
2355
+ "step": 391
2356
+ },
2357
+ {
2358
+ "epoch": 2.45,
2359
+ "learning_rate": 1.9130434782608697e-05,
2360
+ "loss": 0.0585,
2361
+ "step": 392
2362
+ },
2363
+ {
2364
+ "epoch": 2.46,
2365
+ "learning_rate": 1.8913043478260868e-05,
2366
+ "loss": 0.1127,
2367
+ "step": 393
2368
+ },
2369
+ {
2370
+ "epoch": 2.46,
2371
+ "learning_rate": 1.8695652173913045e-05,
2372
+ "loss": 0.0667,
2373
+ "step": 394
2374
+ },
2375
+ {
2376
+ "epoch": 2.47,
2377
+ "learning_rate": 1.8478260869565216e-05,
2378
+ "loss": 0.1232,
2379
+ "step": 395
2380
+ },
2381
+ {
2382
+ "epoch": 2.48,
2383
+ "learning_rate": 1.8260869565217393e-05,
2384
+ "loss": 0.0839,
2385
+ "step": 396
2386
+ },
2387
+ {
2388
+ "epoch": 2.48,
2389
+ "learning_rate": 1.8043478260869567e-05,
2390
+ "loss": 0.0348,
2391
+ "step": 397
2392
+ },
2393
+ {
2394
+ "epoch": 2.49,
2395
+ "learning_rate": 1.782608695652174e-05,
2396
+ "loss": 0.0441,
2397
+ "step": 398
2398
+ },
2399
+ {
2400
+ "epoch": 2.49,
2401
+ "learning_rate": 1.7608695652173915e-05,
2402
+ "loss": 0.0862,
2403
+ "step": 399
2404
+ },
2405
+ {
2406
+ "epoch": 2.5,
2407
+ "learning_rate": 1.739130434782609e-05,
2408
+ "loss": 0.0758,
2409
+ "step": 400
2410
+ }
2411
+ ],
2412
+ "logging_steps": 1,
2413
+ "max_steps": 480,
2414
+ "num_train_epochs": 3,
2415
+ "save_steps": 100,
2416
+ "total_flos": 8.459864193775534e+17,
2417
+ "trial_name": null,
2418
+ "trial_params": null
2419
+ }
checkpoint-400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78323211b785cfcc52412469ba4c0d84e0e6dd0b15b04f489e3e6fd4e5901db4
3
+ size 4536
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "__type": "AddedToken",
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "clean_up_tokenization_spaces": false,
11
+ "eos_token": {
12
+ "__type": "AddedToken",
13
+ "content": "</s>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false
18
+ },
19
+ "legacy": false,
20
+ "model_max_length": 1000000000000000019884624838656,
21
+ "pad_token": null,
22
+ "padding_side": "right",
23
+ "sp_model_kwargs": {},
24
+ "tokenizer_class": "LlamaTokenizer",
25
+ "unk_token": {
26
+ "__type": "AddedToken",
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ },
33
+ "use_default_system_prompt": true
34
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78323211b785cfcc52412469ba4c0d84e0e6dd0b15b04f489e3e6fd4e5901db4
3
+ size 4536