emnakamura
commited on
Commit
•
2870c71
1
Parent(s):
1979000
Update README.md
Browse files
README.md
CHANGED
@@ -1,199 +1,116 @@
|
|
1 |
---
|
2 |
library_name: transformers
|
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 |
-
<!-- This should link to a Dataset Card if possible. -->
|
112 |
-
|
113 |
-
[More Information Needed]
|
114 |
-
|
115 |
-
#### Factors
|
116 |
-
|
117 |
-
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
118 |
-
|
119 |
-
[More Information Needed]
|
120 |
-
|
121 |
-
#### Metrics
|
122 |
-
|
123 |
-
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
124 |
-
|
125 |
-
[More Information Needed]
|
126 |
-
|
127 |
-
### Results
|
128 |
-
|
129 |
-
[More Information Needed]
|
130 |
-
|
131 |
-
#### Summary
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
## Model Examination [optional]
|
136 |
-
|
137 |
-
<!-- Relevant interpretability work for the model goes here -->
|
138 |
-
|
139 |
-
[More Information Needed]
|
140 |
-
|
141 |
-
## Environmental Impact
|
142 |
-
|
143 |
-
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
144 |
-
|
145 |
-
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
146 |
-
|
147 |
-
- **Hardware Type:** [More Information Needed]
|
148 |
-
- **Hours used:** [More Information Needed]
|
149 |
-
- **Cloud Provider:** [More Information Needed]
|
150 |
-
- **Compute Region:** [More Information Needed]
|
151 |
-
- **Carbon Emitted:** [More Information Needed]
|
152 |
-
|
153 |
-
## Technical Specifications [optional]
|
154 |
-
|
155 |
-
### Model Architecture and Objective
|
156 |
-
|
157 |
-
[More Information Needed]
|
158 |
-
|
159 |
-
### Compute Infrastructure
|
160 |
-
|
161 |
-
[More Information Needed]
|
162 |
-
|
163 |
-
#### Hardware
|
164 |
-
|
165 |
-
[More Information Needed]
|
166 |
-
|
167 |
-
#### Software
|
168 |
-
|
169 |
-
[More Information Needed]
|
170 |
-
|
171 |
-
## Citation [optional]
|
172 |
-
|
173 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
174 |
-
|
175 |
-
**BibTeX:**
|
176 |
-
|
177 |
-
[More Information Needed]
|
178 |
-
|
179 |
-
**APA:**
|
180 |
-
|
181 |
-
[More Information Needed]
|
182 |
-
|
183 |
-
## Glossary [optional]
|
184 |
-
|
185 |
-
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
186 |
-
|
187 |
-
[More Information Needed]
|
188 |
-
|
189 |
-
## More Information [optional]
|
190 |
-
|
191 |
-
[More Information Needed]
|
192 |
-
|
193 |
-
## Model Card Authors [optional]
|
194 |
-
|
195 |
-
[More Information Needed]
|
196 |
-
|
197 |
-
## Model Card Contact
|
198 |
-
|
199 |
-
[More Information Needed]
|
|
|
1 |
---
|
2 |
library_name: transformers
|
3 |
+
license: apache-2.0
|
4 |
+
base_model:
|
5 |
+
- flammenai/flammen27-mistral-7B
|
6 |
+
datasets:
|
7 |
+
- flammenai/FlameMix-DPO-v1
|
8 |
+
- flammenai/Grill-preprod-v1_chatML
|
9 |
+
- flammenai/Grill-preprod-v2_chatML
|
10 |
---
|
11 |
+
![image/png](https://huggingface.co/flammenai/Mahou-1.0-mistral-7B/resolve/main/mahou1.png)
|
12 |
+
|
13 |
+
# Mahou-1.2a-mistral-7B
|
14 |
+
|
15 |
+
Mahou is our attempt to build a production-ready conversational/roleplay LLM.
|
16 |
+
|
17 |
+
Future versions will be released iteratively and finetuned from flammen.ai conversational data.
|
18 |
+
|
19 |
+
1.2a is rebased and retrained to improve comphresion and coherence.
|
20 |
+
|
21 |
+
### Chat Format
|
22 |
+
|
23 |
+
This model has been trained to use ChatML format.
|
24 |
+
|
25 |
+
```
|
26 |
+
<|im_start|>system
|
27 |
+
{{system}}<|im_end|>
|
28 |
+
<|im_start|>{{char}}
|
29 |
+
{{message}}<|im_end|>
|
30 |
+
<|im_start|>{{user}}
|
31 |
+
{{message}}<|im_end|>
|
32 |
+
```
|
33 |
+
|
34 |
+
# Roleplay Format
|
35 |
+
|
36 |
+
- Speech without quotes.
|
37 |
+
- Actions in `*asterisks*`
|
38 |
+
|
39 |
+
```
|
40 |
+
*leans against wall cooly* so like, i just casted a super strong spell at magician academy today, not gonna lie, felt badass.
|
41 |
+
```
|
42 |
+
|
43 |
+
### ST Settings
|
44 |
+
|
45 |
+
1. Use ChatML for the Context Template.
|
46 |
+
2. Turn on Instruct Mode for ChatML.
|
47 |
+
3. Use the following stopping strings: `["<", "|", "<|", "\n"]`
|
48 |
+
|
49 |
+
### Method
|
50 |
+
|
51 |
+
Finetuned using an A100 on Google Colab.
|
52 |
+
|
53 |
+
[Fine-tune a Mistral-7b model with Direct Preference Optimization](https://towardsdatascience.com/fine-tune-a-mistral-7b-model-with-direct-preference-optimization-708042745aac) - [Maxime Labonne](https://huggingface.co/mlabonne)
|
54 |
+
|
55 |
+
### Configuration
|
56 |
+
|
57 |
+
LoRA, model, and training settings:
|
58 |
+
|
59 |
+
```python
|
60 |
+
# LoRA configuration
|
61 |
+
peft_config = LoraConfig(
|
62 |
+
r=16,
|
63 |
+
lora_alpha=16,
|
64 |
+
lora_dropout=0.05,
|
65 |
+
bias="none",
|
66 |
+
task_type="CAUSAL_LM",
|
67 |
+
target_modules=['k_proj', 'gate_proj', 'v_proj', 'up_proj', 'q_proj', 'o_proj', 'down_proj']
|
68 |
+
)
|
69 |
+
|
70 |
+
# Model to fine-tune
|
71 |
+
model = AutoModelForCausalLM.from_pretrained(
|
72 |
+
model_name,
|
73 |
+
torch_dtype=torch.bfloat16,
|
74 |
+
load_in_4bit=True
|
75 |
+
)
|
76 |
+
model.config.use_cache = False
|
77 |
+
|
78 |
+
# Reference model
|
79 |
+
ref_model = AutoModelForCausalLM.from_pretrained(
|
80 |
+
model_name,
|
81 |
+
torch_dtype=torch.bfloat16,
|
82 |
+
load_in_4bit=True
|
83 |
+
)
|
84 |
+
|
85 |
+
# Training arguments
|
86 |
+
training_args = TrainingArguments(
|
87 |
+
per_device_train_batch_size=4,
|
88 |
+
gradient_accumulation_steps=4,
|
89 |
+
gradient_checkpointing=True,
|
90 |
+
learning_rate=5e-5,
|
91 |
+
lr_scheduler_type="cosine",
|
92 |
+
max_steps=2000,
|
93 |
+
save_strategy="no",
|
94 |
+
logging_steps=1,
|
95 |
+
output_dir=new_model,
|
96 |
+
optim="paged_adamw_32bit",
|
97 |
+
warmup_steps=100,
|
98 |
+
bf16=True,
|
99 |
+
report_to="wandb",
|
100 |
+
)
|
101 |
+
|
102 |
+
# Create DPO trainer
|
103 |
+
dpo_trainer = DPOTrainer(
|
104 |
+
model,
|
105 |
+
ref_model,
|
106 |
+
args=training_args,
|
107 |
+
train_dataset=dataset,
|
108 |
+
tokenizer=tokenizer,
|
109 |
+
peft_config=peft_config,
|
110 |
+
beta=0.1,
|
111 |
+
force_use_ref_model=True
|
112 |
+
)
|
113 |
+
|
114 |
+
# Fine-tune model with DPO
|
115 |
+
dpo_trainer.train()
|
116 |
+
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|