File size: 10,502 Bytes
5cb7ea4
 
db73b94
 
 
 
 
 
 
 
 
 
 
 
 
5cb7ea4
 
 
 
 
 
04d2813
e9da4b9
 
ba9ac72
db73b94
3f6017d
db73b94
 
 
 
 
 
 
 
ba9ac72
 
db73b94
ba9ac72
2a1b572
 
ba9ac72
db73b94
ba9ac72
e9da4b9
04d2813
 
3f6017d
04d2813
9aab0b8
04d2813
9aab0b8
 
04d2813
 
cba0048
04d2813
 
 
 
 
 
 
 
9aab0b8
04d2813
db73b94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e1a91b0
db73b94
857a80b
c22df8d
 
 
857a80b
c22df8d
 
 
857a80b
c22df8d
 
 
857a80b
c22df8d
 
 
7bc28eb
 
 
 
 
 
 
f92245d
7bc28eb
 
 
 
04d2813
ba9ac72
 
db73b94
 
9aab0b8
04d2813
 
 
db73b94
04d2813
 
 
 
 
68237ea
04d2813
 
 
 
 
db73b94
5417824
04d2813
 
 
 
a9e502e
04d2813
e65c203
04d2813
e65c203
04d2813
68237ea
04d2813
 
 
9083910
04d2813
 
 
 
 
 
 
 
 
 
 
0a472e1
 
 
 
04d2813
0a472e1
 
 
 
04d2813
0a472e1
 
 
 
362821c
 
04d2813
a9e502e
0a472e1
04d2813
 
 
0a472e1
 
a9e502e
 
04d2813
 
e65c203
04d2813
 
 
e65c203
04d2813
0a472e1
 
 
 
 
2c34f8d
db73b94
9083910
 
db73b94
0a472e1
 
 
2b43668
 
0a472e1
 
04d2813
0a472e1
 
 
 
 
 
04d2813
9083910
04d2813
 
0a472e1
04d2813
0a472e1
 
 
 
 
 
 
 
04d2813
 
 
 
 
 
c22df8d
0a472e1
04d2813
 
29273b5
0a472e1
 
 
04d2813
 
4ee79f2
0a472e1
04d2813
0a472e1
 
 
0e74b64
0a472e1
 
 
04d2813
 
0a472e1
 
 
 
04d2813
0a472e1
 
04d2813
0a472e1
 
 
05c1834
 
 
 
 
 
 
 
 
 
 
 
04d2813
 
 
 
 
0a472e1
 
 
e65c203
04d2813
0a472e1
 
 
 
 
04d2813
0a472e1
 
 
2b43668
04d2813
 
 
 
2b43668
 
f2a2029
04d2813
 
 
12de7b7
04d2813
 
 
29273b5
04d2813
 
29273b5
 
 
04d2813
 
9083910
 
 
12de7b7
 
04d2813
0a472e1
04d2813
0a472e1
3f6017d
db73b94
 
 
 
ba9ac72
db73b94
 
0a472e1
04d2813
0a472e1
04d2813
 
 
0a472e1
 
04d2813
 
8552218
04d2813
8552218
 
 
 
 
 
 
 
0a472e1
bc97f9c
04d2813
a4f1241
3f6017d
 
e7e1a77
3f6017d
 
ba9ac72
 
5417824
 
 
 
 
 
ba9ac72
1377400
 
 
 
71d600f
3f6017d
e07bd8a
 
ba9ac72
 
 
04d2813
bc97f9c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
# Axolotl

<div align="center">
  <img src="image/axolotl.png" alt="axolotl" width="160">
  <div>
    <p>
      <b>One repo to finetune them all! </b>
    </p>
    <p>
      Go ahead and axolotl questions!!
    </p>
  </div>
</div>

## Axolotl supports

|          | fp16/fp32 | fp16/fp32 w/ lora | 4bit-quant | 4bit-quant w/flash attention | flash attention | xformers attention |
|----------|:----------|:------------------|------------|------------------------------|-----------------|--------------------|
| llama    | βœ…         | βœ…                 | βœ…          | βœ…                            | βœ…               | βœ…                  |
| Pythia   | βœ…         | βœ…                 | ❌          | ❌                            | ❌               | ❓                  |
| cerebras | βœ…         | βœ…                 | ❌          | ❌                            | ❌               | ❓                  |
| mpt      | βœ…         | ❌                 | ❌          | ❌                            | ❌               | ❓                  |


## Quickstart ⚑

**Requirements**: Python 3.9.

```bash
git clone https://github.com/OpenAccess-AI-Collective/axolotl

pip3 install -e .[int4]

accelerate config

# finetune
accelerate launch scripts/finetune.py examples/4bit-lora-7b/config.yml

# inference
accelerate launch scripts/finetune.py examples/4bit-lora-7b/config.yml \
    --inference --lora_model_dir="./llama-7b-lora-int4"
```

## Installation

### Environment

- Docker
  ```bash
  docker run --gpus '"all"' --rm -it winglian/axolotl:main
  ```
  - `winglian/axolotl:dev`: dev branch
  - `winglian/axolotl-runpod:main`: for runpod

- Conda/Pip venv
  1. Install python **3.9**

  2. Install python dependencies with ONE of the following:
      - `pip3 install -e .[int4]` (recommended)
      - `pip3 install -e .[int4_triton]`
      - `pip3 install -e .`

### Dataset

Have dataset(s) in one of the following format (JSONL recommended):

- `alpaca`: instruction; input(optional)
  ```json
  {"instruction": "...", "input": "...", "output": "..."}
  ```
- `sharegpt`: conversations
  ```json
  {"conversations": [{"from": "...", "value": "..."}]}
  ```
- `completion`: raw corpus
  ```json
  {"text": "..."}
  ```

<details>

<summary>See other formats</summary>

- `jeopardy`: question and answer
  ```json
  {"question": "...", "category": "...", "answer": "..."}
  ```
- `oasst`: instruction
  ```json
  {"INSTRUCTION": "...", "RESPONSE": "..."}
  ```
- `gpteacher`: instruction; input(optional)
  ```json
  {"instruction": "...", "input": "...", "response": "..."}
  ```
- `reflection`: instruction with reflect; input(optional)
  ```json
  {"instruction": "...", "input": "...", "output": "...", "reflection": "...", "corrected": "..."}
  ```
- `explainchoice`: question, choices, (solution OR explanation)
  ```json
  {"question": "...", "choices": ["..."], "solution": "...", "explanation": "..."}
  ```
- `concisechoice`: question, choices, (solution OR explanation)
  ```json
  {"question": "...", "choices": ["..."], "solution": "...", "explanation": "..."}
  ```
- `summarizetldr`: article and summary
  ```json
  {"article": "...", "summary": "..."}
  ```

> Have some new format to propose? Check if it's already defined in [data.py](src/axolotl/utils/data.py) in `dev` branch!

</details>

Optionally, download some datasets, see [data/README.md](data/README.md)

### Config

See sample configs in [configs](configs) folder or [examples](examples) for quick start. It is recommended to duplicate and modify to your needs. The most important options are:

- model
  ```yaml
  base_model: ./llama-7b-hf # local or huggingface repo
  ```
  Note: The code will load the right architecture.

- dataset
  ```yaml
  datasets:
    - path: vicgalle/alpaca-gpt4 # local or huggingface repo
      type: alpaca # format from earlier
  sequence_len: 2048 # max token length / prompt
  ```

- loading
  ```yaml
  load_in_4bit: true
  load_in_8bit: true
  bf16: true # require >=ampere
  fp16: true
  tf32: true # require >=ampere
  ```
  Note: Repo does not do 4-bit quantization.

- lora
  ```yaml
  adapter: lora # qlora or leave blank for full finetune
  lora_r: 8
  lora_alpha: 16
  lora_dropout: 0.05
  lora_target_modules:
    - q_proj
    - v_proj
  ```

<details>

<summary>All yaml options</summary>

```yaml
# this is the huggingface model that contains *.pt, *.safetensors, or *.bin files
# this can also be a relative path to a model on disk
base_model: ./llama-7b-hf
# you can specify an ignore pattern if the model repo contains more than 1 model type (*.pt, etc)
base_model_ignore_patterns:
# if the base_model repo on hf hub doesn't include configuration .json files,
# you can set that here, or leave this empty to default to base_model
base_model_config: ./llama-7b-hf
# If you want to specify the type of model to load, AutoModelForCausalLM is a good choice too
model_type: AutoModelForCausalLM
# Corresponding tokenizer for the model AutoTokenizer is a good choice
tokenizer_type: AutoTokenizer
# Trust remote code for untrusted source
trust_remote_code:

# whether you are training a 4-bit GPTQ quantized model
load_4bit: true
gptq_groupsize: 128 # group size
gptq_model_v1: false # v1 or v2

# this will attempt to quantize the model down to 8 bits and use adam 8 bit optimizer
load_in_8bit: true
# use bitsandbytes 4 bit
load_in_4bit:

# Use CUDA bf16
bf16: true # bool or 'full' for `bf16_full_eval`. require >=ampere
# Use CUDA fp16
fp16: true
# Use CUDA tf32
tf32: true # require >=ampere

# a list of one or more datasets to finetune the model with
datasets:
  # this can be either a hf dataset, or relative path
  - path: vicgalle/alpaca-gpt4
  # The type of prompt to use for training. [alpaca, sharegpt, gpteacher, oasst, reflection]
    type: alpaca # format OR format:prompt_style (chat/instruct)
    data_files: # path to source data files
    shards: # true if use subset data. make sure to set `shards` param also
shards: # number of shards to split dataset into

# axolotl attempts to save the dataset as an arrow after packing the data together so
# subsequent training attempts load faster, relative path
dataset_prepared_path: data/last_run_prepared
# push prepared dataset to hub
push_dataset_to_hub: # repo path
# How much of the dataset to set aside as evaluation. 1 = 100%, 0.50 = 50%, etc
val_set_size: 0.04

# the maximum length of an input to train with, this should typically be less than 2048
# as most models have a token/context limit of 2048
sequence_len: 2048
# max sequence length to concatenate training samples together up to
# inspired by StackLLaMA. see https://huggingface.co/blog/stackllama#supervised-fine-tuning
max_packed_sequence_len: 1024

# if you want to use 'lora' or 'qlora' or leave blank to train all parameters in original model
adapter: lora
# if you already have a lora model trained that you want to load, put that here
# lora hyperparameters
lora_model_dir:
lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
  - q_proj
  - v_proj
#  - k_proj
#  - o_proj
#  - gate_proj
#  - down_proj
#  - up_proj
lora_modules_to_save:
#  - embed_tokens
#  - lm_head
lora_out_dir:
lora_fan_in_fan_out: false

# wandb configuration if you're using it
wandb_mode:
wandb_project:
wandb_watch:
wandb_run_id:
wandb_log_model: # 'checkpoint'

# where to save the finished model to
output_dir: ./completed-model

# training hyperparameters
batch_size: 8
micro_batch_size: 2
eval_batch_size: 2
num_epochs: 3
warmup_steps: 100
learning_rate: 0.00003
logging_steps:

# whether to mask out or include the human's prompt from the training labels
train_on_inputs: false
# don't use this, leads to wonky training (according to someone on the internet)
group_by_length: false

# does not work with current implementation of 4-bit LoRA
gradient_checkpointing: false

# stop training after this many evaluation losses have increased in a row
# https://huggingface.co/transformers/v4.2.2/_modules/transformers/trainer_callback.html#EarlyStoppingCallback
early_stopping_patience: 3

# specify a scheduler and kwargs to use with the optimizer
lr_scheduler: # 'one_cycle' | 'log_sweep' | empty for cosine
lr_scheduler_kwargs:

# for one_cycle optim
lr_div_factor: # learning rate div factor

# for log_sweep optim
log_sweep_min_lr:
log_sweep_max_lr:

# specify optimizer
optimizer:
# specify weight decay
weight_decay:

# whether to use xformers attention patch https://github.com/facebookresearch/xformers:
xformers_attention:
# whether to use flash attention patch https://github.com/HazyResearch/flash-attention:
flash_attention:  # require a100 for llama

# resume from a specific checkpoint dir
resume_from_checkpoint:
# if resume_from_checkpoint isn't set and you simply want it to start where it left off
# be careful with this being turned on between different models
auto_resume_from_checkpoints: false

# don't mess with this, it's here for accelerate and torchrun
local_rank:

# add or change special tokens
special_tokens:
  # bos_token: "<s>"
  # eos_token: "</s>"
  # unk_token: "<unk>"
# add extra tokens
tokens:

# FSDP
fsdp:
fsdp_config:

# Deepspeed
deepspeed:

# Path to torch distx for optim 'adamw_anyprecision'
torchdistx_path:

# Set padding for data collator to 'longest'
collator_pad_to_longest:

# Debug mode
debug:

# Seed
seed:
```

</details>

### Accelerate

Configure accelerate

```bash
accelerate config

# Edit manually
# nano ~/.cache/huggingface/accelerate/default_config.yaml
```

### Train

Run
```bash
accelerate launch scripts/finetune.py configs/your_config.yml
```

### Inference

Pass the appropriate flag to the train command:

- Pretrained LORA:
  ```bash
  --inference --lora_model_dir ./completed-model
  ```
- Full weights finetune:
  ```bash
  --inference --base_model ./completed-model
  ```

### Merge LORA to base

Add below flag to train command above

```bash
--merge_lora --lora_model_dir="./completed-model" --load_in_8bit=False --load_in_4bit=False
```

## Common Errors 🧰

> Cuda out of memory

Please reduce any below
  - `micro_batch_size`
  - `eval_batch_size`
  - `sequence_len`

> RuntimeError: expected scalar type Float but found Half

Try set `fp16: true`

## Need help? πŸ™‹β€β™‚οΈ

Join our [Discord server](https://discord.gg/HhrNrHJPRb) where we can help you

## Contributing 🀝

Bugs? Please check for open issue else create a new [Issue](https://github.com/OpenAccess-AI-Collective/axolotl/issues/new).

PRs are **greatly welcome**!