totally-not-an-llm
commited on
Commit
•
7a2eed5
1
Parent(s):
d77366a
Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +19 -1
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- added_tokens.json +5 -0
- checkpoint-100/README.md +21 -0
- checkpoint-100/adapter_config.json +26 -0
- checkpoint-100/adapter_model.bin +3 -0
- checkpoint-100/adapter_model/README.md +21 -0
- checkpoint-100/adapter_model/adapter_config.json +26 -0
- checkpoint-100/adapter_model/adapter_model.bin +3 -0
- checkpoint-100/optimizer.pt +3 -0
- checkpoint-100/rng_state.pth +3 -0
- checkpoint-100/scheduler.pt +3 -0
- checkpoint-100/trainer_state.json +619 -0
- checkpoint-100/training_args.bin +3 -0
- checkpoint-41/README.md +21 -0
- checkpoint-41/adapter_config.json +26 -0
- checkpoint-41/adapter_model.bin +3 -0
- checkpoint-41/adapter_model/README.md +21 -0
- checkpoint-41/adapter_model/adapter_config.json +26 -0
- checkpoint-41/adapter_model/adapter_model.bin +3 -0
- checkpoint-41/optimizer.pt +3 -0
- checkpoint-41/rng_state.pth +3 -0
- checkpoint-41/scheduler.pt +3 -0
- checkpoint-41/trainer_state.json +265 -0
- checkpoint-41/training_args.bin +3 -0
- checkpoint-62/README.md +21 -0
- checkpoint-62/adapter_config.json +26 -0
- checkpoint-62/adapter_model.bin +3 -0
- checkpoint-62/adapter_model/README.md +21 -0
- checkpoint-62/adapter_model/adapter_config.json +26 -0
- checkpoint-62/adapter_model/adapter_model.bin +3 -0
- checkpoint-62/optimizer.pt +3 -0
- checkpoint-62/rng_state.pth +3 -0
- checkpoint-62/scheduler.pt +3 -0
- checkpoint-62/trainer_state.json +391 -0
- checkpoint-62/training_args.bin +3 -0
- checkpoint-83/README.md +21 -0
- checkpoint-83/adapter_config.json +26 -0
- checkpoint-83/adapter_model.bin +3 -0
- checkpoint-83/adapter_model/README.md +21 -0
- checkpoint-83/adapter_model/adapter_config.json +26 -0
- checkpoint-83/adapter_model/adapter_model.bin +3 -0
- checkpoint-83/optimizer.pt +3 -0
- checkpoint-83/rng_state.pth +3 -0
- checkpoint-83/scheduler.pt +3 -0
- checkpoint-83/trainer_state.json +517 -0
- checkpoint-83/training_args.bin +3 -0
- config.json +39 -0
- special_tokens_map.json +6 -0
README.md
CHANGED
@@ -1,3 +1,21 @@
|
|
1 |
---
|
2 |
-
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 0.6.0.dev0
|
adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:11213a0c00bc5430a832af35fa25de9d9592072a55593290454c5487e99e5a21
|
3 |
+
size 500897101
|
added_tokens.json
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s>": 2,
|
3 |
+
"<s>": 1,
|
4 |
+
"<unk>": 0
|
5 |
+
}
|
checkpoint-100/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 0.6.0.dev0
|
checkpoint-100/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-100/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:11213a0c00bc5430a832af35fa25de9d9592072a55593290454c5487e99e5a21
|
3 |
+
size 500897101
|
checkpoint-100/adapter_model/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 0.6.0.dev0
|
checkpoint-100/adapter_model/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-100/adapter_model/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:11213a0c00bc5430a832af35fa25de9d9592072a55593290454c5487e99e5a21
|
3 |
+
size 500897101
|
checkpoint-100/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1c02d23e0bd4a6cef73a32ebf8402a8b1e1a8f18cbebb1526b512a32ee37be5d
|
3 |
+
size 1001736445
|
checkpoint-100/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c465c2771dab9ba6f712e86e9b71c6011a3dc6f2893016c97be18f355357c6b5
|
3 |
+
size 14575
|
checkpoint-100/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ef30a43d20754fbeee89803ef955a46ca7e92fb8c40af3bdae2eaff291dcf94
|
3 |
+
size 627
|
checkpoint-100/trainer_state.json
ADDED
@@ -0,0 +1,619 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 4.8192771084337345,
|
5 |
+
"eval_steps": 150,
|
6 |
+
"global_step": 100,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.05,
|
13 |
+
"learning_rate": 2.9999999999999997e-05,
|
14 |
+
"loss": 1.2314,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.1,
|
19 |
+
"learning_rate": 5.9999999999999995e-05,
|
20 |
+
"loss": 1.4026,
|
21 |
+
"step": 2
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.14,
|
25 |
+
"learning_rate": 8.999999999999999e-05,
|
26 |
+
"loss": 1.8531,
|
27 |
+
"step": 3
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.19,
|
31 |
+
"learning_rate": 0.00011999999999999999,
|
32 |
+
"loss": 1.3763,
|
33 |
+
"step": 4
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.24,
|
37 |
+
"learning_rate": 0.00015,
|
38 |
+
"loss": 0.961,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.29,
|
43 |
+
"learning_rate": 0.00017999999999999998,
|
44 |
+
"loss": 1.2005,
|
45 |
+
"step": 6
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.34,
|
49 |
+
"learning_rate": 0.00020999999999999998,
|
50 |
+
"loss": 1.1054,
|
51 |
+
"step": 7
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39,
|
55 |
+
"learning_rate": 0.00023999999999999998,
|
56 |
+
"loss": 1.0114,
|
57 |
+
"step": 8
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.43,
|
61 |
+
"learning_rate": 0.00027,
|
62 |
+
"loss": 1.1386,
|
63 |
+
"step": 9
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.48,
|
67 |
+
"learning_rate": 0.0003,
|
68 |
+
"loss": 0.9856,
|
69 |
+
"step": 10
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.53,
|
73 |
+
"learning_rate": 0.00029990862405286433,
|
74 |
+
"loss": 0.7562,
|
75 |
+
"step": 11
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.58,
|
79 |
+
"learning_rate": 0.0002996346075389736,
|
80 |
+
"loss": 1.0498,
|
81 |
+
"step": 12
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 0.63,
|
85 |
+
"learning_rate": 0.00029917828430524096,
|
86 |
+
"loss": 1.2559,
|
87 |
+
"step": 13
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.67,
|
91 |
+
"learning_rate": 0.0002985402103112355,
|
92 |
+
"loss": 1.0074,
|
93 |
+
"step": 14
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.72,
|
97 |
+
"learning_rate": 0.0002977211629518312,
|
98 |
+
"loss": 1.0239,
|
99 |
+
"step": 15
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.77,
|
103 |
+
"learning_rate": 0.0002967221401100708,
|
104 |
+
"loss": 1.3441,
|
105 |
+
"step": 16
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.82,
|
109 |
+
"learning_rate": 0.0002955443589413994,
|
110 |
+
"loss": 1.1692,
|
111 |
+
"step": 17
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.87,
|
115 |
+
"learning_rate": 0.0002941892543907478,
|
116 |
+
"loss": 1.3764,
|
117 |
+
"step": 18
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 0.92,
|
121 |
+
"learning_rate": 0.00029265847744427303,
|
122 |
+
"loss": 0.9064,
|
123 |
+
"step": 19
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 0.96,
|
127 |
+
"learning_rate": 0.0002909538931178862,
|
128 |
+
"loss": 1.4608,
|
129 |
+
"step": 20
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.01,
|
133 |
+
"learning_rate": 0.0002890775781850181,
|
134 |
+
"loss": 1.0947,
|
135 |
+
"step": 21
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.06,
|
139 |
+
"learning_rate": 0.0002870318186463901,
|
140 |
+
"loss": 1.113,
|
141 |
+
"step": 22
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 1.11,
|
145 |
+
"learning_rate": 0.000284819106944875,
|
146 |
+
"loss": 0.9546,
|
147 |
+
"step": 23
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 1.16,
|
151 |
+
"learning_rate": 0.000282442138928839,
|
152 |
+
"loss": 1.1469,
|
153 |
+
"step": 24
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 1.2,
|
157 |
+
"learning_rate": 0.0002799038105676658,
|
158 |
+
"loss": 1.3102,
|
159 |
+
"step": 25
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 1.25,
|
163 |
+
"learning_rate": 0.00027720721442346387,
|
164 |
+
"loss": 1.4763,
|
165 |
+
"step": 26
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 1.3,
|
169 |
+
"learning_rate": 0.0002743556358832562,
|
170 |
+
"loss": 1.0281,
|
171 |
+
"step": 27
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.35,
|
175 |
+
"learning_rate": 0.0002713525491562421,
|
176 |
+
"loss": 1.0988,
|
177 |
+
"step": 28
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.4,
|
181 |
+
"learning_rate": 0.00026820161304100823,
|
182 |
+
"loss": 1.1848,
|
183 |
+
"step": 29
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 1.45,
|
187 |
+
"learning_rate": 0.00026490666646784665,
|
188 |
+
"loss": 0.9901,
|
189 |
+
"step": 30
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 1.49,
|
193 |
+
"learning_rate": 0.00026147172382160914,
|
194 |
+
"loss": 0.8676,
|
195 |
+
"step": 31
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 1.54,
|
199 |
+
"learning_rate": 0.00025790097005079764,
|
200 |
+
"loss": 1.0521,
|
201 |
+
"step": 32
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 1.59,
|
205 |
+
"learning_rate": 0.0002541987555688496,
|
206 |
+
"loss": 1.0436,
|
207 |
+
"step": 33
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 1.64,
|
211 |
+
"learning_rate": 0.0002503695909538287,
|
212 |
+
"loss": 0.9511,
|
213 |
+
"step": 34
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.69,
|
217 |
+
"learning_rate": 0.0002464181414529809,
|
218 |
+
"loss": 1.2208,
|
219 |
+
"step": 35
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.73,
|
223 |
+
"learning_rate": 0.0002423492212988487,
|
224 |
+
"loss": 1.1128,
|
225 |
+
"step": 36
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 1.78,
|
229 |
+
"learning_rate": 0.00023816778784387094,
|
230 |
+
"loss": 1.114,
|
231 |
+
"step": 37
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 1.83,
|
235 |
+
"learning_rate": 0.00023387893552061199,
|
236 |
+
"loss": 0.9318,
|
237 |
+
"step": 38
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 1.88,
|
241 |
+
"learning_rate": 0.0002294878896349807,
|
242 |
+
"loss": 1.6424,
|
243 |
+
"step": 39
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 1.93,
|
247 |
+
"learning_rate": 0.000225,
|
248 |
+
"loss": 0.7759,
|
249 |
+
"step": 40
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 1.98,
|
253 |
+
"learning_rate": 0.00022042073441788358,
|
254 |
+
"loss": 0.8061,
|
255 |
+
"step": 41
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 2.02,
|
259 |
+
"learning_rate": 0.0002157556720183616,
|
260 |
+
"loss": 1.2972,
|
261 |
+
"step": 42
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.07,
|
265 |
+
"learning_rate": 0.00021101049646137003,
|
266 |
+
"loss": 0.9694,
|
267 |
+
"step": 43
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 2.12,
|
271 |
+
"learning_rate": 0.0002061909890123868,
|
272 |
+
"loss": 0.7951,
|
273 |
+
"step": 44
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 2.17,
|
277 |
+
"learning_rate": 0.00020130302149885031,
|
278 |
+
"loss": 0.8049,
|
279 |
+
"step": 45
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 2.22,
|
283 |
+
"learning_rate": 0.0001963525491562421,
|
284 |
+
"loss": 0.7308,
|
285 |
+
"step": 46
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 2.27,
|
289 |
+
"learning_rate": 0.00019134560337254986,
|
290 |
+
"loss": 1.4495,
|
291 |
+
"step": 47
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 2.31,
|
295 |
+
"learning_rate": 0.00018628828433995013,
|
296 |
+
"loss": 0.8338,
|
297 |
+
"step": 48
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 2.36,
|
301 |
+
"learning_rate": 0.00018118675362266385,
|
302 |
+
"loss": 0.8706,
|
303 |
+
"step": 49
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.41,
|
307 |
+
"learning_rate": 0.00017604722665003956,
|
308 |
+
"loss": 1.3767,
|
309 |
+
"step": 50
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 2.46,
|
313 |
+
"learning_rate": 0.0001708759651440098,
|
314 |
+
"loss": 1.5032,
|
315 |
+
"step": 51
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 2.51,
|
319 |
+
"learning_rate": 0.000165679269490148,
|
320 |
+
"loss": 1.3356,
|
321 |
+
"step": 52
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 2.55,
|
325 |
+
"learning_rate": 0.00016046347106161876,
|
326 |
+
"loss": 1.0869,
|
327 |
+
"step": 53
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 2.6,
|
331 |
+
"learning_rate": 0.00015523492450537517,
|
332 |
+
"loss": 0.8374,
|
333 |
+
"step": 54
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 2.65,
|
337 |
+
"learning_rate": 0.00015,
|
338 |
+
"loss": 0.9144,
|
339 |
+
"step": 55
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 2.7,
|
343 |
+
"learning_rate": 0.0001447650754946249,
|
344 |
+
"loss": 0.7246,
|
345 |
+
"step": 56
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.75,
|
349 |
+
"learning_rate": 0.00013953652893838119,
|
350 |
+
"loss": 1.0432,
|
351 |
+
"step": 57
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 2.8,
|
355 |
+
"learning_rate": 0.000134320730509852,
|
356 |
+
"loss": 0.9271,
|
357 |
+
"step": 58
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 2.84,
|
361 |
+
"learning_rate": 0.0001291240348559902,
|
362 |
+
"loss": 0.7968,
|
363 |
+
"step": 59
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 2.89,
|
367 |
+
"learning_rate": 0.00012395277334996044,
|
368 |
+
"loss": 0.9131,
|
369 |
+
"step": 60
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 2.94,
|
373 |
+
"learning_rate": 0.00011881324637733611,
|
374 |
+
"loss": 0.937,
|
375 |
+
"step": 61
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 2.99,
|
379 |
+
"learning_rate": 0.00011371171566004985,
|
380 |
+
"loss": 0.8228,
|
381 |
+
"step": 62
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 3.04,
|
385 |
+
"learning_rate": 0.00010865439662745013,
|
386 |
+
"loss": 1.0797,
|
387 |
+
"step": 63
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.08,
|
391 |
+
"learning_rate": 0.0001036474508437579,
|
392 |
+
"loss": 0.8893,
|
393 |
+
"step": 64
|
394 |
+
},
|
395 |
+
{
|
396 |
+
"epoch": 3.13,
|
397 |
+
"learning_rate": 9.869697850114969e-05,
|
398 |
+
"loss": 0.5448,
|
399 |
+
"step": 65
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 3.18,
|
403 |
+
"learning_rate": 9.380901098761319e-05,
|
404 |
+
"loss": 0.6197,
|
405 |
+
"step": 66
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 3.23,
|
409 |
+
"learning_rate": 8.898950353862998e-05,
|
410 |
+
"loss": 0.6925,
|
411 |
+
"step": 67
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 3.28,
|
415 |
+
"learning_rate": 8.424432798163836e-05,
|
416 |
+
"loss": 0.6065,
|
417 |
+
"step": 68
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 3.33,
|
421 |
+
"learning_rate": 7.957926558211642e-05,
|
422 |
+
"loss": 0.9739,
|
423 |
+
"step": 69
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 3.37,
|
427 |
+
"learning_rate": 7.500000000000002e-05,
|
428 |
+
"loss": 0.8073,
|
429 |
+
"step": 70
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.42,
|
433 |
+
"learning_rate": 7.051211036501928e-05,
|
434 |
+
"loss": 1.5069,
|
435 |
+
"step": 71
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 3.47,
|
439 |
+
"learning_rate": 6.612106447938799e-05,
|
440 |
+
"loss": 0.5149,
|
441 |
+
"step": 72
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 3.52,
|
445 |
+
"learning_rate": 6.183221215612904e-05,
|
446 |
+
"loss": 0.8116,
|
447 |
+
"step": 73
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 3.57,
|
451 |
+
"learning_rate": 5.765077870115125e-05,
|
452 |
+
"loss": 0.5185,
|
453 |
+
"step": 74
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 3.61,
|
457 |
+
"learning_rate": 5.358185854701909e-05,
|
458 |
+
"loss": 0.8031,
|
459 |
+
"step": 75
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 3.66,
|
463 |
+
"learning_rate": 4.963040904617131e-05,
|
464 |
+
"loss": 0.5611,
|
465 |
+
"step": 76
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 3.71,
|
469 |
+
"learning_rate": 4.5801244431150394e-05,
|
470 |
+
"loss": 0.7082,
|
471 |
+
"step": 77
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 3.76,
|
475 |
+
"learning_rate": 4.209902994920235e-05,
|
476 |
+
"loss": 0.8057,
|
477 |
+
"step": 78
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 3.81,
|
481 |
+
"learning_rate": 3.852827617839084e-05,
|
482 |
+
"loss": 1.8981,
|
483 |
+
"step": 79
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 3.86,
|
487 |
+
"learning_rate": 3.509333353215331e-05,
|
488 |
+
"loss": 0.8772,
|
489 |
+
"step": 80
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 3.9,
|
493 |
+
"learning_rate": 3.1798386958991714e-05,
|
494 |
+
"loss": 1.1716,
|
495 |
+
"step": 81
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 3.95,
|
499 |
+
"learning_rate": 2.8647450843757897e-05,
|
500 |
+
"loss": 0.949,
|
501 |
+
"step": 82
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 4.0,
|
505 |
+
"learning_rate": 2.5644364116743755e-05,
|
506 |
+
"loss": 0.901,
|
507 |
+
"step": 83
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 4.05,
|
511 |
+
"learning_rate": 2.2792785576536105e-05,
|
512 |
+
"loss": 1.0778,
|
513 |
+
"step": 84
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 4.1,
|
517 |
+
"learning_rate": 2.009618943233419e-05,
|
518 |
+
"loss": 1.0495,
|
519 |
+
"step": 85
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 4.14,
|
523 |
+
"learning_rate": 1.755786107116095e-05,
|
524 |
+
"loss": 1.1405,
|
525 |
+
"step": 86
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 4.19,
|
529 |
+
"learning_rate": 1.5180893055124977e-05,
|
530 |
+
"loss": 0.6457,
|
531 |
+
"step": 87
|
532 |
+
},
|
533 |
+
{
|
534 |
+
"epoch": 4.24,
|
535 |
+
"learning_rate": 1.2968181353609852e-05,
|
536 |
+
"loss": 0.7194,
|
537 |
+
"step": 88
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 4.29,
|
541 |
+
"learning_rate": 1.0922421814981901e-05,
|
542 |
+
"loss": 0.768,
|
543 |
+
"step": 89
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 4.34,
|
547 |
+
"learning_rate": 9.046106882113751e-06,
|
548 |
+
"loss": 0.7015,
|
549 |
+
"step": 90
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 4.39,
|
553 |
+
"learning_rate": 7.34152255572697e-06,
|
554 |
+
"loss": 0.9132,
|
555 |
+
"step": 91
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 4.43,
|
559 |
+
"learning_rate": 5.810745609252165e-06,
|
560 |
+
"loss": 0.703,
|
561 |
+
"step": 92
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 4.48,
|
565 |
+
"learning_rate": 4.455641058600528e-06,
|
566 |
+
"loss": 0.8964,
|
567 |
+
"step": 93
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"epoch": 4.53,
|
571 |
+
"learning_rate": 3.2778598899291465e-06,
|
572 |
+
"loss": 0.9267,
|
573 |
+
"step": 94
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 4.58,
|
577 |
+
"learning_rate": 2.2788370481687965e-06,
|
578 |
+
"loss": 0.7872,
|
579 |
+
"step": 95
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 4.63,
|
583 |
+
"learning_rate": 1.4597896887644456e-06,
|
584 |
+
"loss": 0.7648,
|
585 |
+
"step": 96
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 4.67,
|
589 |
+
"learning_rate": 8.217156947590064e-07,
|
590 |
+
"loss": 0.5883,
|
591 |
+
"step": 97
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 4.72,
|
595 |
+
"learning_rate": 3.653924610263703e-07,
|
596 |
+
"loss": 0.6093,
|
597 |
+
"step": 98
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 4.77,
|
601 |
+
"learning_rate": 9.137594713563568e-08,
|
602 |
+
"loss": 0.8796,
|
603 |
+
"step": 99
|
604 |
+
},
|
605 |
+
{
|
606 |
+
"epoch": 4.82,
|
607 |
+
"learning_rate": 0.0,
|
608 |
+
"loss": 0.5744,
|
609 |
+
"step": 100
|
610 |
+
}
|
611 |
+
],
|
612 |
+
"logging_steps": 1,
|
613 |
+
"max_steps": 100,
|
614 |
+
"num_train_epochs": 5,
|
615 |
+
"save_steps": 500,
|
616 |
+
"total_flos": 2.55142096011264e+17,
|
617 |
+
"trial_name": null,
|
618 |
+
"trial_params": null
|
619 |
+
}
|
checkpoint-100/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a7871ddb870dca149c8bd176c33a0174c54ce7f51f1623c6ad82c38e78485c96
|
3 |
+
size 4411
|
checkpoint-41/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 0.6.0.dev0
|
checkpoint-41/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-41/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:68ddbd889686556508110f942061578825177da74856785e15f97c63d4f4b005
|
3 |
+
size 500897101
|
checkpoint-41/adapter_model/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 0.6.0.dev0
|
checkpoint-41/adapter_model/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-41/adapter_model/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:68ddbd889686556508110f942061578825177da74856785e15f97c63d4f4b005
|
3 |
+
size 500897101
|
checkpoint-41/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aaa675f9cb3260fc969d19ad606acdb7a06f4141a35b38cbe2965d83b8d6ff59
|
3 |
+
size 1001736445
|
checkpoint-41/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ccf410588401c344d577bf277e576cb89ffef47ae39cc4edbb9da08ba2fbb438
|
3 |
+
size 14575
|
checkpoint-41/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:145b10f6d4a353b24514b0dab0a1f00a25175216b6926714cd4f2a5bf613b00b
|
3 |
+
size 627
|
checkpoint-41/trainer_state.json
ADDED
@@ -0,0 +1,265 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.9759036144578315,
|
5 |
+
"eval_steps": 150,
|
6 |
+
"global_step": 41,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.05,
|
13 |
+
"learning_rate": 2.9999999999999997e-05,
|
14 |
+
"loss": 1.2314,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.1,
|
19 |
+
"learning_rate": 5.9999999999999995e-05,
|
20 |
+
"loss": 1.4026,
|
21 |
+
"step": 2
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.14,
|
25 |
+
"learning_rate": 8.999999999999999e-05,
|
26 |
+
"loss": 1.8531,
|
27 |
+
"step": 3
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.19,
|
31 |
+
"learning_rate": 0.00011999999999999999,
|
32 |
+
"loss": 1.3763,
|
33 |
+
"step": 4
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.24,
|
37 |
+
"learning_rate": 0.00015,
|
38 |
+
"loss": 0.961,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.29,
|
43 |
+
"learning_rate": 0.00017999999999999998,
|
44 |
+
"loss": 1.2005,
|
45 |
+
"step": 6
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.34,
|
49 |
+
"learning_rate": 0.00020999999999999998,
|
50 |
+
"loss": 1.1054,
|
51 |
+
"step": 7
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39,
|
55 |
+
"learning_rate": 0.00023999999999999998,
|
56 |
+
"loss": 1.0114,
|
57 |
+
"step": 8
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.43,
|
61 |
+
"learning_rate": 0.00027,
|
62 |
+
"loss": 1.1386,
|
63 |
+
"step": 9
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.48,
|
67 |
+
"learning_rate": 0.0003,
|
68 |
+
"loss": 0.9856,
|
69 |
+
"step": 10
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.53,
|
73 |
+
"learning_rate": 0.00029990862405286433,
|
74 |
+
"loss": 0.7562,
|
75 |
+
"step": 11
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.58,
|
79 |
+
"learning_rate": 0.0002996346075389736,
|
80 |
+
"loss": 1.0498,
|
81 |
+
"step": 12
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 0.63,
|
85 |
+
"learning_rate": 0.00029917828430524096,
|
86 |
+
"loss": 1.2559,
|
87 |
+
"step": 13
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.67,
|
91 |
+
"learning_rate": 0.0002985402103112355,
|
92 |
+
"loss": 1.0074,
|
93 |
+
"step": 14
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.72,
|
97 |
+
"learning_rate": 0.0002977211629518312,
|
98 |
+
"loss": 1.0239,
|
99 |
+
"step": 15
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.77,
|
103 |
+
"learning_rate": 0.0002967221401100708,
|
104 |
+
"loss": 1.3441,
|
105 |
+
"step": 16
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.82,
|
109 |
+
"learning_rate": 0.0002955443589413994,
|
110 |
+
"loss": 1.1692,
|
111 |
+
"step": 17
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.87,
|
115 |
+
"learning_rate": 0.0002941892543907478,
|
116 |
+
"loss": 1.3764,
|
117 |
+
"step": 18
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 0.92,
|
121 |
+
"learning_rate": 0.00029265847744427303,
|
122 |
+
"loss": 0.9064,
|
123 |
+
"step": 19
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 0.96,
|
127 |
+
"learning_rate": 0.0002909538931178862,
|
128 |
+
"loss": 1.4608,
|
129 |
+
"step": 20
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.01,
|
133 |
+
"learning_rate": 0.0002890775781850181,
|
134 |
+
"loss": 1.0947,
|
135 |
+
"step": 21
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.06,
|
139 |
+
"learning_rate": 0.0002870318186463901,
|
140 |
+
"loss": 1.113,
|
141 |
+
"step": 22
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 1.11,
|
145 |
+
"learning_rate": 0.000284819106944875,
|
146 |
+
"loss": 0.9546,
|
147 |
+
"step": 23
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 1.16,
|
151 |
+
"learning_rate": 0.000282442138928839,
|
152 |
+
"loss": 1.1469,
|
153 |
+
"step": 24
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 1.2,
|
157 |
+
"learning_rate": 0.0002799038105676658,
|
158 |
+
"loss": 1.3102,
|
159 |
+
"step": 25
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 1.25,
|
163 |
+
"learning_rate": 0.00027720721442346387,
|
164 |
+
"loss": 1.4763,
|
165 |
+
"step": 26
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 1.3,
|
169 |
+
"learning_rate": 0.0002743556358832562,
|
170 |
+
"loss": 1.0281,
|
171 |
+
"step": 27
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.35,
|
175 |
+
"learning_rate": 0.0002713525491562421,
|
176 |
+
"loss": 1.0988,
|
177 |
+
"step": 28
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.4,
|
181 |
+
"learning_rate": 0.00026820161304100823,
|
182 |
+
"loss": 1.1848,
|
183 |
+
"step": 29
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 1.45,
|
187 |
+
"learning_rate": 0.00026490666646784665,
|
188 |
+
"loss": 0.9901,
|
189 |
+
"step": 30
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 1.49,
|
193 |
+
"learning_rate": 0.00026147172382160914,
|
194 |
+
"loss": 0.8676,
|
195 |
+
"step": 31
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 1.54,
|
199 |
+
"learning_rate": 0.00025790097005079764,
|
200 |
+
"loss": 1.0521,
|
201 |
+
"step": 32
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 1.59,
|
205 |
+
"learning_rate": 0.0002541987555688496,
|
206 |
+
"loss": 1.0436,
|
207 |
+
"step": 33
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 1.64,
|
211 |
+
"learning_rate": 0.0002503695909538287,
|
212 |
+
"loss": 0.9511,
|
213 |
+
"step": 34
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.69,
|
217 |
+
"learning_rate": 0.0002464181414529809,
|
218 |
+
"loss": 1.2208,
|
219 |
+
"step": 35
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.73,
|
223 |
+
"learning_rate": 0.0002423492212988487,
|
224 |
+
"loss": 1.1128,
|
225 |
+
"step": 36
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 1.78,
|
229 |
+
"learning_rate": 0.00023816778784387094,
|
230 |
+
"loss": 1.114,
|
231 |
+
"step": 37
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 1.83,
|
235 |
+
"learning_rate": 0.00023387893552061199,
|
236 |
+
"loss": 0.9318,
|
237 |
+
"step": 38
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 1.88,
|
241 |
+
"learning_rate": 0.0002294878896349807,
|
242 |
+
"loss": 1.6424,
|
243 |
+
"step": 39
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 1.93,
|
247 |
+
"learning_rate": 0.000225,
|
248 |
+
"loss": 0.7759,
|
249 |
+
"step": 40
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 1.98,
|
253 |
+
"learning_rate": 0.00022042073441788358,
|
254 |
+
"loss": 0.8061,
|
255 |
+
"step": 41
|
256 |
+
}
|
257 |
+
],
|
258 |
+
"logging_steps": 1,
|
259 |
+
"max_steps": 100,
|
260 |
+
"num_train_epochs": 5,
|
261 |
+
"save_steps": 500,
|
262 |
+
"total_flos": 1.0588396984467456e+17,
|
263 |
+
"trial_name": null,
|
264 |
+
"trial_params": null
|
265 |
+
}
|
checkpoint-41/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a7871ddb870dca149c8bd176c33a0174c54ce7f51f1623c6ad82c38e78485c96
|
3 |
+
size 4411
|
checkpoint-62/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 0.6.0.dev0
|
checkpoint-62/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-62/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3996fa6cd634cef2e886ba53704cdaf585f63488956dc493d6208f5bd88b5f89
|
3 |
+
size 500897101
|
checkpoint-62/adapter_model/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 0.6.0.dev0
|
checkpoint-62/adapter_model/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-62/adapter_model/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3996fa6cd634cef2e886ba53704cdaf585f63488956dc493d6208f5bd88b5f89
|
3 |
+
size 500897101
|
checkpoint-62/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8b0ad741144f35cb085b4e4b6f6090ac1ee2d46ee5d3b4c75e1c88dde85c72d1
|
3 |
+
size 1001736445
|
checkpoint-62/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:819629affa360efc5b4f9241c8241641e09fe79d70e6c056cd560592ddce2847
|
3 |
+
size 14575
|
checkpoint-62/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:885848ada3813b8df84a9c8eb6f0a75ba666603a30e5d56b09dfd1fbca05085f
|
3 |
+
size 627
|
checkpoint-62/trainer_state.json
ADDED
@@ -0,0 +1,391 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.9879518072289155,
|
5 |
+
"eval_steps": 150,
|
6 |
+
"global_step": 62,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.05,
|
13 |
+
"learning_rate": 2.9999999999999997e-05,
|
14 |
+
"loss": 1.2314,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.1,
|
19 |
+
"learning_rate": 5.9999999999999995e-05,
|
20 |
+
"loss": 1.4026,
|
21 |
+
"step": 2
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.14,
|
25 |
+
"learning_rate": 8.999999999999999e-05,
|
26 |
+
"loss": 1.8531,
|
27 |
+
"step": 3
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.19,
|
31 |
+
"learning_rate": 0.00011999999999999999,
|
32 |
+
"loss": 1.3763,
|
33 |
+
"step": 4
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.24,
|
37 |
+
"learning_rate": 0.00015,
|
38 |
+
"loss": 0.961,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.29,
|
43 |
+
"learning_rate": 0.00017999999999999998,
|
44 |
+
"loss": 1.2005,
|
45 |
+
"step": 6
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.34,
|
49 |
+
"learning_rate": 0.00020999999999999998,
|
50 |
+
"loss": 1.1054,
|
51 |
+
"step": 7
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39,
|
55 |
+
"learning_rate": 0.00023999999999999998,
|
56 |
+
"loss": 1.0114,
|
57 |
+
"step": 8
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.43,
|
61 |
+
"learning_rate": 0.00027,
|
62 |
+
"loss": 1.1386,
|
63 |
+
"step": 9
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.48,
|
67 |
+
"learning_rate": 0.0003,
|
68 |
+
"loss": 0.9856,
|
69 |
+
"step": 10
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.53,
|
73 |
+
"learning_rate": 0.00029990862405286433,
|
74 |
+
"loss": 0.7562,
|
75 |
+
"step": 11
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.58,
|
79 |
+
"learning_rate": 0.0002996346075389736,
|
80 |
+
"loss": 1.0498,
|
81 |
+
"step": 12
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 0.63,
|
85 |
+
"learning_rate": 0.00029917828430524096,
|
86 |
+
"loss": 1.2559,
|
87 |
+
"step": 13
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.67,
|
91 |
+
"learning_rate": 0.0002985402103112355,
|
92 |
+
"loss": 1.0074,
|
93 |
+
"step": 14
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.72,
|
97 |
+
"learning_rate": 0.0002977211629518312,
|
98 |
+
"loss": 1.0239,
|
99 |
+
"step": 15
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.77,
|
103 |
+
"learning_rate": 0.0002967221401100708,
|
104 |
+
"loss": 1.3441,
|
105 |
+
"step": 16
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.82,
|
109 |
+
"learning_rate": 0.0002955443589413994,
|
110 |
+
"loss": 1.1692,
|
111 |
+
"step": 17
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.87,
|
115 |
+
"learning_rate": 0.0002941892543907478,
|
116 |
+
"loss": 1.3764,
|
117 |
+
"step": 18
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 0.92,
|
121 |
+
"learning_rate": 0.00029265847744427303,
|
122 |
+
"loss": 0.9064,
|
123 |
+
"step": 19
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 0.96,
|
127 |
+
"learning_rate": 0.0002909538931178862,
|
128 |
+
"loss": 1.4608,
|
129 |
+
"step": 20
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.01,
|
133 |
+
"learning_rate": 0.0002890775781850181,
|
134 |
+
"loss": 1.0947,
|
135 |
+
"step": 21
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.06,
|
139 |
+
"learning_rate": 0.0002870318186463901,
|
140 |
+
"loss": 1.113,
|
141 |
+
"step": 22
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 1.11,
|
145 |
+
"learning_rate": 0.000284819106944875,
|
146 |
+
"loss": 0.9546,
|
147 |
+
"step": 23
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 1.16,
|
151 |
+
"learning_rate": 0.000282442138928839,
|
152 |
+
"loss": 1.1469,
|
153 |
+
"step": 24
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 1.2,
|
157 |
+
"learning_rate": 0.0002799038105676658,
|
158 |
+
"loss": 1.3102,
|
159 |
+
"step": 25
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 1.25,
|
163 |
+
"learning_rate": 0.00027720721442346387,
|
164 |
+
"loss": 1.4763,
|
165 |
+
"step": 26
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 1.3,
|
169 |
+
"learning_rate": 0.0002743556358832562,
|
170 |
+
"loss": 1.0281,
|
171 |
+
"step": 27
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.35,
|
175 |
+
"learning_rate": 0.0002713525491562421,
|
176 |
+
"loss": 1.0988,
|
177 |
+
"step": 28
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.4,
|
181 |
+
"learning_rate": 0.00026820161304100823,
|
182 |
+
"loss": 1.1848,
|
183 |
+
"step": 29
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 1.45,
|
187 |
+
"learning_rate": 0.00026490666646784665,
|
188 |
+
"loss": 0.9901,
|
189 |
+
"step": 30
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 1.49,
|
193 |
+
"learning_rate": 0.00026147172382160914,
|
194 |
+
"loss": 0.8676,
|
195 |
+
"step": 31
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 1.54,
|
199 |
+
"learning_rate": 0.00025790097005079764,
|
200 |
+
"loss": 1.0521,
|
201 |
+
"step": 32
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 1.59,
|
205 |
+
"learning_rate": 0.0002541987555688496,
|
206 |
+
"loss": 1.0436,
|
207 |
+
"step": 33
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 1.64,
|
211 |
+
"learning_rate": 0.0002503695909538287,
|
212 |
+
"loss": 0.9511,
|
213 |
+
"step": 34
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.69,
|
217 |
+
"learning_rate": 0.0002464181414529809,
|
218 |
+
"loss": 1.2208,
|
219 |
+
"step": 35
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.73,
|
223 |
+
"learning_rate": 0.0002423492212988487,
|
224 |
+
"loss": 1.1128,
|
225 |
+
"step": 36
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 1.78,
|
229 |
+
"learning_rate": 0.00023816778784387094,
|
230 |
+
"loss": 1.114,
|
231 |
+
"step": 37
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 1.83,
|
235 |
+
"learning_rate": 0.00023387893552061199,
|
236 |
+
"loss": 0.9318,
|
237 |
+
"step": 38
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 1.88,
|
241 |
+
"learning_rate": 0.0002294878896349807,
|
242 |
+
"loss": 1.6424,
|
243 |
+
"step": 39
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 1.93,
|
247 |
+
"learning_rate": 0.000225,
|
248 |
+
"loss": 0.7759,
|
249 |
+
"step": 40
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 1.98,
|
253 |
+
"learning_rate": 0.00022042073441788358,
|
254 |
+
"loss": 0.8061,
|
255 |
+
"step": 41
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 2.02,
|
259 |
+
"learning_rate": 0.0002157556720183616,
|
260 |
+
"loss": 1.2972,
|
261 |
+
"step": 42
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.07,
|
265 |
+
"learning_rate": 0.00021101049646137003,
|
266 |
+
"loss": 0.9694,
|
267 |
+
"step": 43
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 2.12,
|
271 |
+
"learning_rate": 0.0002061909890123868,
|
272 |
+
"loss": 0.7951,
|
273 |
+
"step": 44
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 2.17,
|
277 |
+
"learning_rate": 0.00020130302149885031,
|
278 |
+
"loss": 0.8049,
|
279 |
+
"step": 45
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 2.22,
|
283 |
+
"learning_rate": 0.0001963525491562421,
|
284 |
+
"loss": 0.7308,
|
285 |
+
"step": 46
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 2.27,
|
289 |
+
"learning_rate": 0.00019134560337254986,
|
290 |
+
"loss": 1.4495,
|
291 |
+
"step": 47
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 2.31,
|
295 |
+
"learning_rate": 0.00018628828433995013,
|
296 |
+
"loss": 0.8338,
|
297 |
+
"step": 48
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 2.36,
|
301 |
+
"learning_rate": 0.00018118675362266385,
|
302 |
+
"loss": 0.8706,
|
303 |
+
"step": 49
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.41,
|
307 |
+
"learning_rate": 0.00017604722665003956,
|
308 |
+
"loss": 1.3767,
|
309 |
+
"step": 50
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 2.46,
|
313 |
+
"learning_rate": 0.0001708759651440098,
|
314 |
+
"loss": 1.5032,
|
315 |
+
"step": 51
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 2.51,
|
319 |
+
"learning_rate": 0.000165679269490148,
|
320 |
+
"loss": 1.3356,
|
321 |
+
"step": 52
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 2.55,
|
325 |
+
"learning_rate": 0.00016046347106161876,
|
326 |
+
"loss": 1.0869,
|
327 |
+
"step": 53
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 2.6,
|
331 |
+
"learning_rate": 0.00015523492450537517,
|
332 |
+
"loss": 0.8374,
|
333 |
+
"step": 54
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 2.65,
|
337 |
+
"learning_rate": 0.00015,
|
338 |
+
"loss": 0.9144,
|
339 |
+
"step": 55
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 2.7,
|
343 |
+
"learning_rate": 0.0001447650754946249,
|
344 |
+
"loss": 0.7246,
|
345 |
+
"step": 56
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.75,
|
349 |
+
"learning_rate": 0.00013953652893838119,
|
350 |
+
"loss": 1.0432,
|
351 |
+
"step": 57
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 2.8,
|
355 |
+
"learning_rate": 0.000134320730509852,
|
356 |
+
"loss": 0.9271,
|
357 |
+
"step": 58
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 2.84,
|
361 |
+
"learning_rate": 0.0001291240348559902,
|
362 |
+
"loss": 0.7968,
|
363 |
+
"step": 59
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 2.89,
|
367 |
+
"learning_rate": 0.00012395277334996044,
|
368 |
+
"loss": 0.9131,
|
369 |
+
"step": 60
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 2.94,
|
373 |
+
"learning_rate": 0.00011881324637733611,
|
374 |
+
"loss": 0.937,
|
375 |
+
"step": 61
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 2.99,
|
379 |
+
"learning_rate": 0.00011371171566004985,
|
380 |
+
"loss": 0.8228,
|
381 |
+
"step": 62
|
382 |
+
}
|
383 |
+
],
|
384 |
+
"logging_steps": 1,
|
385 |
+
"max_steps": 100,
|
386 |
+
"num_train_epochs": 5,
|
387 |
+
"save_steps": 500,
|
388 |
+
"total_flos": 1.5882595476701184e+17,
|
389 |
+
"trial_name": null,
|
390 |
+
"trial_params": null
|
391 |
+
}
|
checkpoint-62/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a7871ddb870dca149c8bd176c33a0174c54ce7f51f1623c6ad82c38e78485c96
|
3 |
+
size 4411
|
checkpoint-83/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 0.6.0.dev0
|
checkpoint-83/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-83/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:879cb7c9f89b873d886fb62e7cdcf8a9440aab1e33335fe885cb7d5a49ae78ae
|
3 |
+
size 500897101
|
checkpoint-83/adapter_model/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 0.6.0.dev0
|
checkpoint-83/adapter_model/adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": null,
|
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": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"k_proj",
|
19 |
+
"o_proj",
|
20 |
+
"gate_proj",
|
21 |
+
"up_proj",
|
22 |
+
"v_proj",
|
23 |
+
"down_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
checkpoint-83/adapter_model/adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:879cb7c9f89b873d886fb62e7cdcf8a9440aab1e33335fe885cb7d5a49ae78ae
|
3 |
+
size 500897101
|
checkpoint-83/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bd942b91e8b634407667729822b11d68150258cb2c2c428268928f0a1dcc7c83
|
3 |
+
size 1001736445
|
checkpoint-83/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e54251fd1f9b73f220ddc492ed2e7d2b0f6cd2f7c70974412a6eccd366a58c77
|
3 |
+
size 14575
|
checkpoint-83/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:877f8c8498f1154c5fce2ccfa856cadaf35802ef3e318bde01bf77c6f3f136b9
|
3 |
+
size 627
|
checkpoint-83/trainer_state.json
ADDED
@@ -0,0 +1,517 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 4.0,
|
5 |
+
"eval_steps": 150,
|
6 |
+
"global_step": 83,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.05,
|
13 |
+
"learning_rate": 2.9999999999999997e-05,
|
14 |
+
"loss": 1.2314,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.1,
|
19 |
+
"learning_rate": 5.9999999999999995e-05,
|
20 |
+
"loss": 1.4026,
|
21 |
+
"step": 2
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.14,
|
25 |
+
"learning_rate": 8.999999999999999e-05,
|
26 |
+
"loss": 1.8531,
|
27 |
+
"step": 3
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.19,
|
31 |
+
"learning_rate": 0.00011999999999999999,
|
32 |
+
"loss": 1.3763,
|
33 |
+
"step": 4
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.24,
|
37 |
+
"learning_rate": 0.00015,
|
38 |
+
"loss": 0.961,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.29,
|
43 |
+
"learning_rate": 0.00017999999999999998,
|
44 |
+
"loss": 1.2005,
|
45 |
+
"step": 6
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.34,
|
49 |
+
"learning_rate": 0.00020999999999999998,
|
50 |
+
"loss": 1.1054,
|
51 |
+
"step": 7
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39,
|
55 |
+
"learning_rate": 0.00023999999999999998,
|
56 |
+
"loss": 1.0114,
|
57 |
+
"step": 8
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.43,
|
61 |
+
"learning_rate": 0.00027,
|
62 |
+
"loss": 1.1386,
|
63 |
+
"step": 9
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.48,
|
67 |
+
"learning_rate": 0.0003,
|
68 |
+
"loss": 0.9856,
|
69 |
+
"step": 10
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.53,
|
73 |
+
"learning_rate": 0.00029990862405286433,
|
74 |
+
"loss": 0.7562,
|
75 |
+
"step": 11
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.58,
|
79 |
+
"learning_rate": 0.0002996346075389736,
|
80 |
+
"loss": 1.0498,
|
81 |
+
"step": 12
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 0.63,
|
85 |
+
"learning_rate": 0.00029917828430524096,
|
86 |
+
"loss": 1.2559,
|
87 |
+
"step": 13
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.67,
|
91 |
+
"learning_rate": 0.0002985402103112355,
|
92 |
+
"loss": 1.0074,
|
93 |
+
"step": 14
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.72,
|
97 |
+
"learning_rate": 0.0002977211629518312,
|
98 |
+
"loss": 1.0239,
|
99 |
+
"step": 15
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.77,
|
103 |
+
"learning_rate": 0.0002967221401100708,
|
104 |
+
"loss": 1.3441,
|
105 |
+
"step": 16
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.82,
|
109 |
+
"learning_rate": 0.0002955443589413994,
|
110 |
+
"loss": 1.1692,
|
111 |
+
"step": 17
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.87,
|
115 |
+
"learning_rate": 0.0002941892543907478,
|
116 |
+
"loss": 1.3764,
|
117 |
+
"step": 18
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 0.92,
|
121 |
+
"learning_rate": 0.00029265847744427303,
|
122 |
+
"loss": 0.9064,
|
123 |
+
"step": 19
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 0.96,
|
127 |
+
"learning_rate": 0.0002909538931178862,
|
128 |
+
"loss": 1.4608,
|
129 |
+
"step": 20
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.01,
|
133 |
+
"learning_rate": 0.0002890775781850181,
|
134 |
+
"loss": 1.0947,
|
135 |
+
"step": 21
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.06,
|
139 |
+
"learning_rate": 0.0002870318186463901,
|
140 |
+
"loss": 1.113,
|
141 |
+
"step": 22
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 1.11,
|
145 |
+
"learning_rate": 0.000284819106944875,
|
146 |
+
"loss": 0.9546,
|
147 |
+
"step": 23
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 1.16,
|
151 |
+
"learning_rate": 0.000282442138928839,
|
152 |
+
"loss": 1.1469,
|
153 |
+
"step": 24
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 1.2,
|
157 |
+
"learning_rate": 0.0002799038105676658,
|
158 |
+
"loss": 1.3102,
|
159 |
+
"step": 25
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 1.25,
|
163 |
+
"learning_rate": 0.00027720721442346387,
|
164 |
+
"loss": 1.4763,
|
165 |
+
"step": 26
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 1.3,
|
169 |
+
"learning_rate": 0.0002743556358832562,
|
170 |
+
"loss": 1.0281,
|
171 |
+
"step": 27
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.35,
|
175 |
+
"learning_rate": 0.0002713525491562421,
|
176 |
+
"loss": 1.0988,
|
177 |
+
"step": 28
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.4,
|
181 |
+
"learning_rate": 0.00026820161304100823,
|
182 |
+
"loss": 1.1848,
|
183 |
+
"step": 29
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 1.45,
|
187 |
+
"learning_rate": 0.00026490666646784665,
|
188 |
+
"loss": 0.9901,
|
189 |
+
"step": 30
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 1.49,
|
193 |
+
"learning_rate": 0.00026147172382160914,
|
194 |
+
"loss": 0.8676,
|
195 |
+
"step": 31
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 1.54,
|
199 |
+
"learning_rate": 0.00025790097005079764,
|
200 |
+
"loss": 1.0521,
|
201 |
+
"step": 32
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 1.59,
|
205 |
+
"learning_rate": 0.0002541987555688496,
|
206 |
+
"loss": 1.0436,
|
207 |
+
"step": 33
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 1.64,
|
211 |
+
"learning_rate": 0.0002503695909538287,
|
212 |
+
"loss": 0.9511,
|
213 |
+
"step": 34
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.69,
|
217 |
+
"learning_rate": 0.0002464181414529809,
|
218 |
+
"loss": 1.2208,
|
219 |
+
"step": 35
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.73,
|
223 |
+
"learning_rate": 0.0002423492212988487,
|
224 |
+
"loss": 1.1128,
|
225 |
+
"step": 36
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 1.78,
|
229 |
+
"learning_rate": 0.00023816778784387094,
|
230 |
+
"loss": 1.114,
|
231 |
+
"step": 37
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 1.83,
|
235 |
+
"learning_rate": 0.00023387893552061199,
|
236 |
+
"loss": 0.9318,
|
237 |
+
"step": 38
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 1.88,
|
241 |
+
"learning_rate": 0.0002294878896349807,
|
242 |
+
"loss": 1.6424,
|
243 |
+
"step": 39
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 1.93,
|
247 |
+
"learning_rate": 0.000225,
|
248 |
+
"loss": 0.7759,
|
249 |
+
"step": 40
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 1.98,
|
253 |
+
"learning_rate": 0.00022042073441788358,
|
254 |
+
"loss": 0.8061,
|
255 |
+
"step": 41
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 2.02,
|
259 |
+
"learning_rate": 0.0002157556720183616,
|
260 |
+
"loss": 1.2972,
|
261 |
+
"step": 42
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.07,
|
265 |
+
"learning_rate": 0.00021101049646137003,
|
266 |
+
"loss": 0.9694,
|
267 |
+
"step": 43
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 2.12,
|
271 |
+
"learning_rate": 0.0002061909890123868,
|
272 |
+
"loss": 0.7951,
|
273 |
+
"step": 44
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 2.17,
|
277 |
+
"learning_rate": 0.00020130302149885031,
|
278 |
+
"loss": 0.8049,
|
279 |
+
"step": 45
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 2.22,
|
283 |
+
"learning_rate": 0.0001963525491562421,
|
284 |
+
"loss": 0.7308,
|
285 |
+
"step": 46
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 2.27,
|
289 |
+
"learning_rate": 0.00019134560337254986,
|
290 |
+
"loss": 1.4495,
|
291 |
+
"step": 47
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 2.31,
|
295 |
+
"learning_rate": 0.00018628828433995013,
|
296 |
+
"loss": 0.8338,
|
297 |
+
"step": 48
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 2.36,
|
301 |
+
"learning_rate": 0.00018118675362266385,
|
302 |
+
"loss": 0.8706,
|
303 |
+
"step": 49
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.41,
|
307 |
+
"learning_rate": 0.00017604722665003956,
|
308 |
+
"loss": 1.3767,
|
309 |
+
"step": 50
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 2.46,
|
313 |
+
"learning_rate": 0.0001708759651440098,
|
314 |
+
"loss": 1.5032,
|
315 |
+
"step": 51
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 2.51,
|
319 |
+
"learning_rate": 0.000165679269490148,
|
320 |
+
"loss": 1.3356,
|
321 |
+
"step": 52
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 2.55,
|
325 |
+
"learning_rate": 0.00016046347106161876,
|
326 |
+
"loss": 1.0869,
|
327 |
+
"step": 53
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 2.6,
|
331 |
+
"learning_rate": 0.00015523492450537517,
|
332 |
+
"loss": 0.8374,
|
333 |
+
"step": 54
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 2.65,
|
337 |
+
"learning_rate": 0.00015,
|
338 |
+
"loss": 0.9144,
|
339 |
+
"step": 55
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 2.7,
|
343 |
+
"learning_rate": 0.0001447650754946249,
|
344 |
+
"loss": 0.7246,
|
345 |
+
"step": 56
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.75,
|
349 |
+
"learning_rate": 0.00013953652893838119,
|
350 |
+
"loss": 1.0432,
|
351 |
+
"step": 57
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 2.8,
|
355 |
+
"learning_rate": 0.000134320730509852,
|
356 |
+
"loss": 0.9271,
|
357 |
+
"step": 58
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 2.84,
|
361 |
+
"learning_rate": 0.0001291240348559902,
|
362 |
+
"loss": 0.7968,
|
363 |
+
"step": 59
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 2.89,
|
367 |
+
"learning_rate": 0.00012395277334996044,
|
368 |
+
"loss": 0.9131,
|
369 |
+
"step": 60
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 2.94,
|
373 |
+
"learning_rate": 0.00011881324637733611,
|
374 |
+
"loss": 0.937,
|
375 |
+
"step": 61
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 2.99,
|
379 |
+
"learning_rate": 0.00011371171566004985,
|
380 |
+
"loss": 0.8228,
|
381 |
+
"step": 62
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 3.04,
|
385 |
+
"learning_rate": 0.00010865439662745013,
|
386 |
+
"loss": 1.0797,
|
387 |
+
"step": 63
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.08,
|
391 |
+
"learning_rate": 0.0001036474508437579,
|
392 |
+
"loss": 0.8893,
|
393 |
+
"step": 64
|
394 |
+
},
|
395 |
+
{
|
396 |
+
"epoch": 3.13,
|
397 |
+
"learning_rate": 9.869697850114969e-05,
|
398 |
+
"loss": 0.5448,
|
399 |
+
"step": 65
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 3.18,
|
403 |
+
"learning_rate": 9.380901098761319e-05,
|
404 |
+
"loss": 0.6197,
|
405 |
+
"step": 66
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 3.23,
|
409 |
+
"learning_rate": 8.898950353862998e-05,
|
410 |
+
"loss": 0.6925,
|
411 |
+
"step": 67
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 3.28,
|
415 |
+
"learning_rate": 8.424432798163836e-05,
|
416 |
+
"loss": 0.6065,
|
417 |
+
"step": 68
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 3.33,
|
421 |
+
"learning_rate": 7.957926558211642e-05,
|
422 |
+
"loss": 0.9739,
|
423 |
+
"step": 69
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 3.37,
|
427 |
+
"learning_rate": 7.500000000000002e-05,
|
428 |
+
"loss": 0.8073,
|
429 |
+
"step": 70
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.42,
|
433 |
+
"learning_rate": 7.051211036501928e-05,
|
434 |
+
"loss": 1.5069,
|
435 |
+
"step": 71
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 3.47,
|
439 |
+
"learning_rate": 6.612106447938799e-05,
|
440 |
+
"loss": 0.5149,
|
441 |
+
"step": 72
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 3.52,
|
445 |
+
"learning_rate": 6.183221215612904e-05,
|
446 |
+
"loss": 0.8116,
|
447 |
+
"step": 73
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 3.57,
|
451 |
+
"learning_rate": 5.765077870115125e-05,
|
452 |
+
"loss": 0.5185,
|
453 |
+
"step": 74
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 3.61,
|
457 |
+
"learning_rate": 5.358185854701909e-05,
|
458 |
+
"loss": 0.8031,
|
459 |
+
"step": 75
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 3.66,
|
463 |
+
"learning_rate": 4.963040904617131e-05,
|
464 |
+
"loss": 0.5611,
|
465 |
+
"step": 76
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 3.71,
|
469 |
+
"learning_rate": 4.5801244431150394e-05,
|
470 |
+
"loss": 0.7082,
|
471 |
+
"step": 77
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 3.76,
|
475 |
+
"learning_rate": 4.209902994920235e-05,
|
476 |
+
"loss": 0.8057,
|
477 |
+
"step": 78
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 3.81,
|
481 |
+
"learning_rate": 3.852827617839084e-05,
|
482 |
+
"loss": 1.8981,
|
483 |
+
"step": 79
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 3.86,
|
487 |
+
"learning_rate": 3.509333353215331e-05,
|
488 |
+
"loss": 0.8772,
|
489 |
+
"step": 80
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 3.9,
|
493 |
+
"learning_rate": 3.1798386958991714e-05,
|
494 |
+
"loss": 1.1716,
|
495 |
+
"step": 81
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 3.95,
|
499 |
+
"learning_rate": 2.8647450843757897e-05,
|
500 |
+
"loss": 0.949,
|
501 |
+
"step": 82
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 4.0,
|
505 |
+
"learning_rate": 2.5644364116743755e-05,
|
506 |
+
"loss": 0.901,
|
507 |
+
"step": 83
|
508 |
+
}
|
509 |
+
],
|
510 |
+
"logging_steps": 1,
|
511 |
+
"max_steps": 100,
|
512 |
+
"num_train_epochs": 5,
|
513 |
+
"save_steps": 500,
|
514 |
+
"total_flos": 2.1176793968934912e+17,
|
515 |
+
"trial_name": null,
|
516 |
+
"trial_params": null
|
517 |
+
}
|
checkpoint-83/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a7871ddb870dca149c8bd176c33a0174c54ce7f51f1623c6ad82c38e78485c96
|
3 |
+
size 4411
|
config.json
ADDED
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "NousResearch/Llama-2-13b-hf",
|
3 |
+
"architectures": [
|
4 |
+
"LlamaForCausalLM"
|
5 |
+
],
|
6 |
+
"bos_token_id": 1,
|
7 |
+
"eos_token_id": 2,
|
8 |
+
"hidden_act": "silu",
|
9 |
+
"hidden_size": 5120,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 13824,
|
12 |
+
"max_position_embeddings": 4096,
|
13 |
+
"model_type": "llama",
|
14 |
+
"num_attention_heads": 40,
|
15 |
+
"num_hidden_layers": 40,
|
16 |
+
"num_key_value_heads": 40,
|
17 |
+
"pad_token_id": 0,
|
18 |
+
"pretraining_tp": 1,
|
19 |
+
"quantization_config": {
|
20 |
+
"bnb_4bit_compute_dtype": "bfloat16",
|
21 |
+
"bnb_4bit_quant_type": "nf4",
|
22 |
+
"bnb_4bit_use_double_quant": true,
|
23 |
+
"llm_int8_enable_fp32_cpu_offload": false,
|
24 |
+
"llm_int8_has_fp16_weight": false,
|
25 |
+
"llm_int8_skip_modules": null,
|
26 |
+
"llm_int8_threshold": 6.0,
|
27 |
+
"load_in_4bit": true,
|
28 |
+
"load_in_8bit": false,
|
29 |
+
"quant_method": "bitsandbytes"
|
30 |
+
},
|
31 |
+
"rms_norm_eps": 1e-05,
|
32 |
+
"rope_scaling": null,
|
33 |
+
"rope_theta": 10000.0,
|
34 |
+
"tie_word_embeddings": false,
|
35 |
+
"torch_dtype": "float16",
|
36 |
+
"transformers_version": "4.34.0.dev0",
|
37 |
+
"use_cache": false,
|
38 |
+
"vocab_size": 32000
|
39 |
+
}
|
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 |
+
}
|