sharpenb commited on
Commit
488f8bb
1 Parent(s): 287824f

Upload folder using huggingface_hub (#3)

Browse files

- 75393391d393c1aeb6f04418d5a708d4c43c251e79c7221a94d6bfb661fa150c (a2821f419ba1ab08c6812426573c0fc4017d8f19)

Files changed (4) hide show
  1. README.md +2 -2
  2. config.json +2 -2
  3. plots.png +0 -0
  4. smash_config.json +1 -1
README.md CHANGED
@@ -34,7 +34,7 @@ tags:
34
 
35
  ## Results
36
 
37
- Detailed efficiency metrics coming soon!
38
 
39
  **Frequently Asked Questions**
40
  - ***How does the compression work?*** The model is compressed with llm-int8.
@@ -61,7 +61,7 @@ You can run the smashed model with these steps:
61
  from transformers import AutoModelForCausalLM, AutoTokenizer
62
 
63
  model = AutoModelForCausalLM.from_pretrained("PrunaAI/TinyLlama-TinyLlama-1.1B-Chat-v0.3-bnb-8bit-smashed",
64
- trust_remote_code=True)
65
  tokenizer = AutoTokenizer.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v0.3")
66
 
67
  input_ids = tokenizer("What is the color of prunes?,", return_tensors='pt').to(model.device)["input_ids"]
 
34
 
35
  ## Results
36
 
37
+ ![image info](./plots.png)
38
 
39
  **Frequently Asked Questions**
40
  - ***How does the compression work?*** The model is compressed with llm-int8.
 
61
  from transformers import AutoModelForCausalLM, AutoTokenizer
62
 
63
  model = AutoModelForCausalLM.from_pretrained("PrunaAI/TinyLlama-TinyLlama-1.1B-Chat-v0.3-bnb-8bit-smashed",
64
+ trust_remote_code=True, device_map='auto')
65
  tokenizer = AutoTokenizer.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v0.3")
66
 
67
  input_ids = tokenizer("What is the color of prunes?,", return_tensors='pt').to(model.device)["input_ids"]
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/tmp/tmplj_ek5qh",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
@@ -20,7 +20,7 @@
20
  "quantization_config": {
21
  "bnb_4bit_compute_dtype": "bfloat16",
22
  "bnb_4bit_quant_type": "fp4",
23
- "bnb_4bit_use_double_quant": true,
24
  "llm_int8_enable_fp32_cpu_offload": false,
25
  "llm_int8_has_fp16_weight": false,
26
  "llm_int8_skip_modules": [
 
1
  {
2
+ "_name_or_path": "/tmp/tmpmhdqs5ju",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
 
20
  "quantization_config": {
21
  "bnb_4bit_compute_dtype": "bfloat16",
22
  "bnb_4bit_quant_type": "fp4",
23
+ "bnb_4bit_use_double_quant": false,
24
  "llm_int8_enable_fp32_cpu_offload": false,
25
  "llm_int8_has_fp16_weight": false,
26
  "llm_int8_skip_modules": [
plots.png ADDED
smash_config.json CHANGED
@@ -8,7 +8,7 @@
8
  "compilers": "None",
9
  "task": "text_text_generation",
10
  "device": "cuda",
11
- "cache_dir": "/ceph/hdd/staff/charpent/.cache/models768aaa6l",
12
  "batch_size": 1,
13
  "model_name": "TinyLlama/TinyLlama-1.1B-Chat-v0.3",
14
  "pruning_ratio": 0.0,
 
8
  "compilers": "None",
9
  "task": "text_text_generation",
10
  "device": "cuda",
11
+ "cache_dir": "/ceph/hdd/staff/charpent/.cache/modelsfd34wod3",
12
  "batch_size": 1,
13
  "model_name": "TinyLlama/TinyLlama-1.1B-Chat-v0.3",
14
  "pruning_ratio": 0.0,