Upload 6 files
Browse files- README.md +80 -0
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- training_log.json +16 -0
- training_parameters.json +30 -0
- training_prompt.json +5 -0
README.md
ADDED
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
---
|
4 |
+
## Training procedure
|
5 |
+
|
6 |
+
|
7 |
+
The following `bitsandbytes` quantization config was used during training:
|
8 |
+
- load_in_8bit: False
|
9 |
+
- load_in_4bit: True
|
10 |
+
- llm_int8_threshold: 6.0
|
11 |
+
- llm_int8_skip_modules: None
|
12 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
13 |
+
- llm_int8_has_fp16_weight: False
|
14 |
+
- bnb_4bit_quant_type: nf4
|
15 |
+
- bnb_4bit_use_double_quant: False
|
16 |
+
- bnb_4bit_compute_dtype: bfloat16
|
17 |
+
|
18 |
+
The following `bitsandbytes` quantization config was used during training:
|
19 |
+
- load_in_8bit: False
|
20 |
+
- load_in_4bit: True
|
21 |
+
- llm_int8_threshold: 6.0
|
22 |
+
- llm_int8_skip_modules: None
|
23 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
24 |
+
- llm_int8_has_fp16_weight: False
|
25 |
+
- bnb_4bit_quant_type: nf4
|
26 |
+
- bnb_4bit_use_double_quant: False
|
27 |
+
- bnb_4bit_compute_dtype: bfloat16
|
28 |
+
|
29 |
+
The following `bitsandbytes` quantization config was used during training:
|
30 |
+
- load_in_8bit: False
|
31 |
+
- load_in_4bit: True
|
32 |
+
- llm_int8_threshold: 6.0
|
33 |
+
- llm_int8_skip_modules: None
|
34 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
35 |
+
- llm_int8_has_fp16_weight: False
|
36 |
+
- bnb_4bit_quant_type: nf4
|
37 |
+
- bnb_4bit_use_double_quant: False
|
38 |
+
- bnb_4bit_compute_dtype: bfloat16
|
39 |
+
|
40 |
+
The following `bitsandbytes` quantization config was used during training:
|
41 |
+
- load_in_8bit: False
|
42 |
+
- load_in_4bit: True
|
43 |
+
- llm_int8_threshold: 6.0
|
44 |
+
- llm_int8_skip_modules: None
|
45 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
46 |
+
- llm_int8_has_fp16_weight: False
|
47 |
+
- bnb_4bit_quant_type: nf4
|
48 |
+
- bnb_4bit_use_double_quant: False
|
49 |
+
- bnb_4bit_compute_dtype: bfloat16
|
50 |
+
|
51 |
+
The following `bitsandbytes` quantization config was used during training:
|
52 |
+
- load_in_8bit: False
|
53 |
+
- load_in_4bit: True
|
54 |
+
- llm_int8_threshold: 6.0
|
55 |
+
- llm_int8_skip_modules: None
|
56 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
57 |
+
- llm_int8_has_fp16_weight: False
|
58 |
+
- bnb_4bit_quant_type: nf4
|
59 |
+
- bnb_4bit_use_double_quant: False
|
60 |
+
- bnb_4bit_compute_dtype: bfloat16
|
61 |
+
|
62 |
+
The following `bitsandbytes` quantization config was used during training:
|
63 |
+
- load_in_8bit: False
|
64 |
+
- load_in_4bit: True
|
65 |
+
- llm_int8_threshold: 6.0
|
66 |
+
- llm_int8_skip_modules: None
|
67 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
68 |
+
- llm_int8_has_fp16_weight: False
|
69 |
+
- bnb_4bit_quant_type: nf4
|
70 |
+
- bnb_4bit_use_double_quant: False
|
71 |
+
- bnb_4bit_compute_dtype: bfloat16
|
72 |
+
### Framework versions
|
73 |
+
|
74 |
+
- PEFT 0.5.0.dev0
|
75 |
+
- PEFT 0.5.0.dev0
|
76 |
+
- PEFT 0.5.0.dev0
|
77 |
+
- PEFT 0.5.0.dev0
|
78 |
+
- PEFT 0.5.0.dev0
|
79 |
+
|
80 |
+
- PEFT 0.5.0.dev0
|
adapter_config.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "models\\LogicLMv3",
|
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": 128,
|
11 |
+
"lora_dropout": 0.05,
|
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:a90f44885bcfaa1c73b781bbfdba32715cfc4a664a886750b2582bbe7f4a5848
|
3 |
+
size 209772877
|
training_log.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"base_model_name": "LogicLMv3",
|
3 |
+
"base_model_class": "LlamaForCausalLM",
|
4 |
+
"base_loaded_in_4bit": true,
|
5 |
+
"base_loaded_in_8bit": false,
|
6 |
+
"projections": "q, v",
|
7 |
+
"loss": 1.1674,
|
8 |
+
"learning_rate": 2.359882005899705e-07,
|
9 |
+
"epoch": 3.99,
|
10 |
+
"current_steps": 21979,
|
11 |
+
"train_runtime": 258780.2694,
|
12 |
+
"train_samples_per_second": 1.023,
|
13 |
+
"train_steps_per_second": 0.001,
|
14 |
+
"total_flos": 2.6619410217015706e+18,
|
15 |
+
"train_loss": 1.2536086592563362
|
16 |
+
}
|
training_parameters.json
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"lora_name": "wwpp",
|
3 |
+
"always_override": true,
|
4 |
+
"save_steps": 0.0,
|
5 |
+
"micro_batch_size": 12,
|
6 |
+
"batch_size": 768,
|
7 |
+
"epochs": 4.0,
|
8 |
+
"learning_rate": "2e-5",
|
9 |
+
"lr_scheduler_type": "linear",
|
10 |
+
"lora_rank": 64,
|
11 |
+
"lora_alpha": 128,
|
12 |
+
"lora_dropout": 0.05,
|
13 |
+
"cutoff_len": 256,
|
14 |
+
"dataset": "komt_squad",
|
15 |
+
"eval_dataset": "None",
|
16 |
+
"format": "alpaca-format",
|
17 |
+
"eval_steps": 100.0,
|
18 |
+
"raw_text_file": "None",
|
19 |
+
"overlap_len": 128,
|
20 |
+
"newline_favor_len": 128,
|
21 |
+
"higher_rank_limit": false,
|
22 |
+
"warmup_steps": 300.0,
|
23 |
+
"optimizer": "adamw_torch",
|
24 |
+
"hard_cut_string": "\\f",
|
25 |
+
"train_only_after": "",
|
26 |
+
"stop_at_loss": 0,
|
27 |
+
"add_eos_token": true,
|
28 |
+
"min_chars": null,
|
29 |
+
"report_to": "tensorboard"
|
30 |
+
}
|
training_prompt.json
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"template_type": "dataset",
|
3 |
+
"template_1": "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n%instruction%\n\n### Response:\n%output%",
|
4 |
+
"template_2": "Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n\n### Instruction:\n%instruction%\n\n### Input:\n%input%\n\n### Response:\n%output%"
|
5 |
+
}
|