add adapter
Browse files- README.md +63 -0
- adapter_config.json +28 -0
- adapter_model.bin +3 -0
- info.txt +5 -0
README.md
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
base_model: codellama/CodeLlama-7b-Instruct-hf
|
4 |
+
---
|
5 |
+
|
6 |
+
# Model Card for Model ID
|
7 |
+
|
8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
9 |
+
|
10 |
+
|
11 |
+
|
12 |
+
## Model Details
|
13 |
+
|
14 |
+
### Model Description
|
15 |
+
|
16 |
+
<!-- Provide a longer summary of what this model is. -->
|
17 |
+
|
18 |
+
|
19 |
+
|
20 |
+
- **Developed by:** Riccardo Andrea Izzo
|
21 |
+
- **Model type:** Transformer-based language model
|
22 |
+
- **Language(s) (NLP):** English
|
23 |
+
- **Finetuned from model [optional]:** CodeLlama-7b-Instruct-hf
|
24 |
+
|
25 |
+
### Model Sources [optional]
|
26 |
+
|
27 |
+
<!-- Provide the basic links for the model. -->
|
28 |
+
|
29 |
+
- **Repository:** codellama/CodeLlama-7b-Instruct-hf
|
30 |
+
|
31 |
+
## Uses
|
32 |
+
|
33 |
+
Behavior trees generation for robotic tasks
|
34 |
+
|
35 |
+
## Environmental Impact
|
36 |
+
|
37 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
38 |
+
|
39 |
+
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).
|
40 |
+
|
41 |
+
- **Hardware Type:** 2x NVIDIA Quadro RTX 6000
|
42 |
+
- **Hours used:** 36h
|
43 |
+
|
44 |
+
|
45 |
+
## Training procedure
|
46 |
+
|
47 |
+
|
48 |
+
The following `bitsandbytes` quantization config was used during training:
|
49 |
+
- quant_method: bitsandbytes
|
50 |
+
- load_in_8bit: True
|
51 |
+
- load_in_4bit: False
|
52 |
+
- llm_int8_threshold: 6.0
|
53 |
+
- llm_int8_skip_modules: None
|
54 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
55 |
+
- llm_int8_has_fp16_weight: False
|
56 |
+
- bnb_4bit_quant_type: fp4
|
57 |
+
- bnb_4bit_use_double_quant: False
|
58 |
+
- bnb_4bit_compute_dtype: float32
|
59 |
+
|
60 |
+
### Framework versions
|
61 |
+
|
62 |
+
|
63 |
+
- PEFT 0.6.0
|
adapter_config.json
ADDED
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": null,
|
4 |
+
"base_model_name_or_path": "codellama/CodeLlama-7b-Instruct-hf",
|
5 |
+
"bias": "none",
|
6 |
+
"fan_in_fan_out": false,
|
7 |
+
"inference_mode": true,
|
8 |
+
"init_lora_weights": true,
|
9 |
+
"layers_pattern": null,
|
10 |
+
"layers_to_transform": null,
|
11 |
+
"lora_alpha": 16,
|
12 |
+
"lora_dropout": 0.05,
|
13 |
+
"modules_to_save": null,
|
14 |
+
"peft_type": "LORA",
|
15 |
+
"r": 8,
|
16 |
+
"rank_pattern": {},
|
17 |
+
"revision": null,
|
18 |
+
"target_modules": [
|
19 |
+
"k_proj",
|
20 |
+
"o_proj",
|
21 |
+
"q_proj",
|
22 |
+
"up_proj",
|
23 |
+
"v_proj",
|
24 |
+
"down_proj",
|
25 |
+
"gate_proj"
|
26 |
+
],
|
27 |
+
"task_type": "CAUSAL_LM"
|
28 |
+
}
|
adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:72bb90d4b97ecd19105e15c950bb41ae7967cba486f9d2ce387b01ad55b6aae0
|
3 |
+
size 80114765
|
info.txt
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
--batch_size 256
|
2 |
+
--micro_batch_size 4
|
3 |
+
--cutoff_len 2048
|
4 |
+
|
5 |
+
loss: 0.0982 - epoch: 75.29
|