Instruction LoRA upload — task instruction-lora-test4

README.md

@@ -1,13 +1,5 @@
-# Instruction SFT — Subnet 56
-
-- **Task ID:** instruction-local-test3
+# LoRA — Instruction SFT
+- **Task ID:** instruction-lora-test4
 - **Base model:** mistralai/Mistral-7B-Instruct-v0.2
-- **SHA256 (model.safetensors / adapter_model.safetensors):** `d5d9e891e9ddf087b29ad5127b8ab20beeb8e94ab60209f0f4a2b81a10ca0f24`
+- **SHA256 (adapter):** `dd3b1ad5769464176f93c866adf449537632b4ccaed367436d628673d936f749`
 - **Repo:** raniero/instruction_local_test
-
-## Uso locale (esempio)
-```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
-base = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
-tok = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
-```

adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "peft_type": "LORA",
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
+  "revision": null,
+  "task_type": "CAUSAL_LM",
+  "inference_mode": false,
+  "r": 8,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "fan_in_fan_out": false,
+  "bias": "none",
+  "modules_to_save": null,
+  "init_lora_weights": true,
+  "layers_to_transform": null,
+  "layers_pattern": null,
+  "rank_pattern": {},
+  "alpha_pattern": {},
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "loftq_config": {}
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd3b1ad5769464176f93c866adf449537632b4ccaed367436d628673d936f749
+size 13648400