Upload folder using huggingface_hub (#3)

- 16eb33ee893ebbfacea3c05f9b771000156789bcd5228d158e57873628fccb95 (b2f88c89be918e67d131b973c3ebc416b66ed40f)
- 7657f03b65fa534e57255511cb43b63f60ac673413ea8a8586ead71a2c61bb3d (0d47b8230aad1a306be505089b02e0e47c431746)

README.md

@@ -34,7 +34,7 @@ tags:
 
 ## Results
 
-Detailed efficiency metrics coming soon!
+![image info](./plots.png)
 
 **Frequently Asked Questions**
 - ***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:
    from transformers import AutoModelForCausalLM, AutoTokenizer
 
     model = AutoModelForCausalLM.from_pretrained("PrunaAI/Nondzu-zephyr-speakleash-010-pl-3072-32-16-0.01-bnb-4bit-smashed",
-                                                 trust_remote_code=True)
+                                                 trust_remote_code=True, device_map='auto')
     tokenizer = AutoTokenizer.from_pretrained("Nondzu/zephyr-speakleash-010-pl-3072-32-16-0.01")
     
     input_ids = tokenizer("What is the color of prunes?,", return_tensors='pt').to(model.device)["input_ids"]

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/tmp/tmp8x38bjkc",
+  "_name_or_path": "/tmp/tmpx6gh8mxt",
   "architectures": [
     "MistralForCausalLM"
   ],
@@ -19,7 +19,7 @@
   "quantization_config": {
     "bnb_4bit_compute_dtype": "bfloat16",
     "bnb_4bit_quant_type": "fp4",
-    "bnb_4bit_use_double_quant": true,
+    "bnb_4bit_use_double_quant": false,
     "llm_int8_enable_fp32_cpu_offload": false,
     "llm_int8_has_fp16_weight": false,
     "llm_int8_skip_modules": [

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9fa63ea5a91afa479451e3cd7e2fbf7f77934c335666f6118dfcc17d601d1bab
-size 4125687620
+oid sha256:7f73a5c451e4df14b74c39164224d03bcbf58c442ea768b66e1bb028c389d6fa
+size 4450837496

smash_config.json

@@ -8,7 +8,7 @@
         "compilers": "None",
         "task": "text_text_generation",
         "device": "cuda",
-        "cache_dir": "/ceph/hdd/staff/charpent/.cache/models53xkt2do",
+        "cache_dir": "/ceph/hdd/staff/charpent/.cache/models0xj90l2n",
         "batch_size": 1,
         "model_name": "Nondzu/zephyr-speakleash-010-pl-3072-32-16-0.01",
         "pruning_ratio": 0.0,