llama-anon
commited on
Commit
•
44b0c54
1
Parent(s):
e44eea4
Upload 3 files
Browse files- adapter_config.json +21 -0
- training_log.json +16 -0
- training_parameters.json +30 -0
adapter_config.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-13b-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": 128,
|
11 |
+
"lora_dropout": 0.01,
|
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 |
+
}
|
training_log.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"base_model_name": "Llama2-13B-GPTQ",
|
3 |
+
"base_model_class": "LlamaForCausalLM",
|
4 |
+
"base_loaded_in_4bit": false,
|
5 |
+
"base_loaded_in_8bit": false,
|
6 |
+
"projections": "q, v",
|
7 |
+
"loss": 1.7243,
|
8 |
+
"learning_rate": 6.733428315694157e-07,
|
9 |
+
"epoch": 1.0,
|
10 |
+
"current_steps": 9599,
|
11 |
+
"train_runtime": 3878.3685,
|
12 |
+
"train_samples_per_second": 2.479,
|
13 |
+
"train_steps_per_second": 0.019,
|
14 |
+
"total_flos": 3195128512512000.0,
|
15 |
+
"train_loss": 1.8589664967854818
|
16 |
+
}
|
training_parameters.json
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"lora_name": "lmg2",
|
3 |
+
"always_override": false,
|
4 |
+
"save_steps": 0.0,
|
5 |
+
"micro_batch_size": 1,
|
6 |
+
"batch_size": 128,
|
7 |
+
"epochs": 1.0,
|
8 |
+
"learning_rate": "0.00006",
|
9 |
+
"lr_scheduler_type": "cosine",
|
10 |
+
"lora_rank": 64,
|
11 |
+
"lora_alpha": 128,
|
12 |
+
"lora_dropout": 0.01,
|
13 |
+
"cutoff_len": 256,
|
14 |
+
"dataset": "None",
|
15 |
+
"eval_dataset": "None",
|
16 |
+
"format": "None",
|
17 |
+
"eval_steps": 100.0,
|
18 |
+
"raw_text_file": "lmg2",
|
19 |
+
"overlap_len": 128,
|
20 |
+
"newline_favor_len": 128,
|
21 |
+
"higher_rank_limit": false,
|
22 |
+
"warmup_steps": 100.0,
|
23 |
+
"optimizer": "adamw_torch",
|
24 |
+
"hard_cut_string": "\\n\\n\\n",
|
25 |
+
"train_only_after": "",
|
26 |
+
"stop_at_loss": 0,
|
27 |
+
"add_eos_token": false,
|
28 |
+
"min_chars": 0.0,
|
29 |
+
"report_to": "None"
|
30 |
+
}
|