Create config.json

config.json

@@ -0,0 +1,51 @@
+{
+  "model": {
+    "architecture": "llama",
+    "name": "mistralai_mistral-7b-instruct-v0.1",
+    "model_type": "7B",
+    "quantization": "Q4_K - Medium",
+    "parameters": {
+      "n_vocab": 32000,
+      "n_merges": 0,
+      "n_embd": 4096,
+      "n_ff": 14336,
+      "n_layer": 32,
+      "n_head": 32,
+      "n_head_kv": 8,
+      "n_ctx_train": 32768,
+      "n_embd_head_k": 128,
+      "n_embd_head_v": 128,
+      "n_gqa": 4,
+      "rope_dimension_count": 128,
+      "rope_scaling": "linear",
+      "freq_base_train": 10000.0,
+      "freq_scale_train": 1,
+      "attention_layer_norm_rms_epsilon": 0.00001
+    }
+  },
+  "tokenizer": {
+    "type": "SPM",
+    "vocab_size": 32000,
+    "special_tokens": {
+      "BOS": 1,
+      "EOS": 2,
+      "UNK": 0,
+      "LF": 13
+    }
+  },
+  "file_metadata": {
+    "file_type": "GGUF V2",
+    "model_params": "7.24 B",
+    "model_size": "4.07 GiB",
+    "block_count": 32,
+    "embedding_length": 4096,
+    "quantization_version": 2,
+    "model_path": "Models\\mistral-7b-instruct-v0.2.Q3_K_M.gguf"
+  },
+  "device": "cpu",
+  "notes": {
+    "warning": "WARNING! device is not default parameter. device was transferred to model_kwargs.",
+    "model_loading_details": "Loaded meta data with 20 key-value pairs and 291 tensors.",
+    "special_tokens_check": "Special tokens definition check successful (259/32000)."
+  }
+}