English
Generated from Trainer
Kotokin commited on
Commit
74335cd
1 Parent(s): c532cfb

Upload 7 files

Browse files
adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "models\\HuggingFaceH4_zephyr-7b-beta",
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": 64,
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
+ "v_proj",
20
+ "o_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d76c4123d8e5a6899e02586d9944082008a2c3ba16e3d63f51d1bb19d597b73
3
+ size 109144714
training_graph.json ADDED
The diff for this file is too large to render. See raw diff
 
training_graph.png ADDED
training_log.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name": "HuggingFaceH4_zephyr-7b-beta",
3
+ "base_model_class": "MistralForCausalLM",
4
+ "base_loaded_in_4bit": true,
5
+ "base_loaded_in_8bit": false,
6
+ "projections": "q, k, v, o",
7
+ "loss": 2.3056,
8
+ "learning_rate": 0.0003,
9
+ "epoch": 1.0,
10
+ "current_steps": 2144,
11
+ "current_steps_adjusted": 2144,
12
+ "epoch_adjusted": 1.0,
13
+ "train_runtime": 25358.375,
14
+ "train_samples_per_second": 4.228,
15
+ "train_steps_per_second": 0.085,
16
+ "total_flos": 1.175467232744964e+18,
17
+ "train_loss": 2.3212102342040946
18
+ }
training_parameters.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "lora_name": "NSFW1",
3
+ "always_override": true,
4
+ "save_steps": 0.0,
5
+ "micro_batch_size": 50,
6
+ "batch_size": 0,
7
+ "epochs": 1.0,
8
+ "learning_rate": "3e-4",
9
+ "lr_scheduler_type": "constant",
10
+ "lora_rank": 32,
11
+ "lora_alpha": 64,
12
+ "lora_dropout": 0.05,
13
+ "cutoff_len": 256,
14
+ "dataset": "None",
15
+ "eval_dataset": "None",
16
+ "format": "None",
17
+ "eval_steps": 100.0,
18
+ "raw_text_file": "newfile",
19
+ "higher_rank_limit": false,
20
+ "warmup_steps": 0.0,
21
+ "optimizer": "adamw_torch",
22
+ "hard_cut_string": "\\n\\n\\n",
23
+ "train_only_after": "",
24
+ "stop_at_loss": 0,
25
+ "add_eos_token": false,
26
+ "min_chars": 0.0,
27
+ "report_to": "None",
28
+ "precize_slicing_overlap": true,
29
+ "add_eos_token_type": "Hard Cut Blocks Only",
30
+ "save_steps_under_loss": 1.2,
31
+ "add_bos_token": false,
32
+ "training_projection": "q-k-v-o",
33
+ "sliding_window": false,
34
+ "warmup_ratio": 0,
35
+ "grad_accumulation": 1,
36
+ "neft_noise_alpha": 0
37
+ }
training_prompt.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "template_type": "raw_text"
3
+ }