Upload model

README.md

@@ -12,8 +12,8 @@ The following `bitsandbytes` quantization config was used during training:
 - llm_int8_enable_fp32_cpu_offload: False
 - llm_int8_has_fp16_weight: False
 - bnb_4bit_quant_type: nf4
-- bnb_4bit_use_double_quant: False
-- bnb_4bit_compute_dtype: float16
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
 ### Framework versions
 
 

adapter_config.json

@@ -1,17 +1,17 @@
 {
   "auto_mapping": null,
-  "base_model_name_or_path": "tiiuae/falcon-7b-instruct",
+  "base_model_name_or_path": "tiiuae/falcon-7b",
   "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": true,
   "init_lora_weights": true,
   "layers_pattern": null,
   "layers_to_transform": null,
-  "lora_alpha": 32,
+  "lora_alpha": 8,
   "lora_dropout": 0.05,
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 16,
+  "r": 2,
   "revision": null,
   "target_modules": [
     "query_key_value"

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5f0ae9852bec0f3a2fc14dd47bef976db0e137b83ed2bcdfc68406d854f7218e
-size 18898161
+oid sha256:3a8d8b3480fed8864e5724a227f9446a6dc393e9c9c71954dfb0a61d5833cf8a
+size 2382961