GPTQ model commit

config.json

@@ -1,7 +1,7 @@
 {
     "_name_or_path": "/workspace/process/undi95_mixtral-8x7b-moe-rp-story/source",
     "architectures": [
-        "MistralForCausalLM"
+        "MixtralForCausalLM"
     ],
     "attention_dropout": 0.0,
     "bos_token_id": 1,
@@ -20,52 +20,73 @@
     "output_router_logits": false,
     "pad_token_id": 0,
     "pretraining_tp": 1,
-    "rms_norm_eps": 1e-05,
-    "rope_theta": 10000.0,
-    "router_aux_loss_coef": 0.001,
-    "sliding_window": 4096,
-    "tie_word_embeddings": false,
-    "torch_dtype": "bfloat16",
-    "transformers_version": "4.37.0.dev0",
-    "use_cache": true,
-    "vocab_size": 32000,
     "quantization_config": {
+        "batch_size": 1,
         "bits": 4,
-	"modules_in_block_to_quantize" :[
-        ["self_attn.k_proj", "self_attn.v_proj", "self_attn.q_proj"],
-        ["self_attn.o_proj"],
-        [
-            "block_sparse_moe.experts.0.w1",
-            "block_sparse_moe.experts.1.w1",
-            "block_sparse_moe.experts.2.w1",
-            "block_sparse_moe.experts.3.w1",
-            "block_sparse_moe.experts.4.w1",
-            "block_sparse_moe.experts.5.w1",
-            "block_sparse_moe.experts.6.w1",
-            "block_sparse_moe.experts.7.w1",
-            "block_sparse_moe.experts.0.w3",
-            "block_sparse_moe.experts.1.w3",
-            "block_sparse_moe.experts.2.w3",
-            "block_sparse_moe.experts.3.w3",
-            "block_sparse_moe.experts.4.w3",
-            "block_sparse_moe.experts.5.w3",
-            "block_sparse_moe.experts.6.w3",
-            "block_sparse_moe.experts.7.w3" ],
-        [ "block_sparse_moe.experts.0.w2",
-            "block_sparse_moe.experts.1.w2",
-            "block_sparse_moe.experts.2.w2",
-            "block_sparse_moe.experts.3.w2",
-            "block_sparse_moe.experts.4.w2",
-            "block_sparse_moe.experts.5.w2",
-            "block_sparse_moe.experts.6.w2",
-            "block_sparse_moe.experts.7.w2" ] ],
-        "group_size": -1,
+        "block_name_to_quantize": null,
+        "cache_block_outputs": true,
         "damp_percent": 0.1,
         "desc_act": true,
+        "exllama_config": {
+            "version": 1
+        },
+        "group_size": -1,
+        "max_input_length": null,
+        "model_seqlen": null,
+        "module_name_preceding_first_block": null,
+        "modules_in_block_to_quantize": [
+            [
+                "self_attn.k_proj",
+                "self_attn.v_proj",
+                "self_attn.q_proj"
+            ],
+            [
+                "self_attn.o_proj"
+            ],
+            [
+                "block_sparse_moe.experts.0.w1",
+                "block_sparse_moe.experts.1.w1",
+                "block_sparse_moe.experts.2.w1",
+                "block_sparse_moe.experts.3.w1",
+                "block_sparse_moe.experts.4.w1",
+                "block_sparse_moe.experts.5.w1",
+                "block_sparse_moe.experts.6.w1",
+                "block_sparse_moe.experts.7.w1",
+                "block_sparse_moe.experts.0.w3",
+                "block_sparse_moe.experts.1.w3",
+                "block_sparse_moe.experts.2.w3",
+                "block_sparse_moe.experts.3.w3",
+                "block_sparse_moe.experts.4.w3",
+                "block_sparse_moe.experts.5.w3",
+                "block_sparse_moe.experts.6.w3",
+                "block_sparse_moe.experts.7.w3"
+            ],
+            [
+                "block_sparse_moe.experts.0.w2",
+                "block_sparse_moe.experts.1.w2",
+                "block_sparse_moe.experts.2.w2",
+                "block_sparse_moe.experts.3.w2",
+                "block_sparse_moe.experts.4.w2",
+                "block_sparse_moe.experts.5.w2",
+                "block_sparse_moe.experts.6.w2",
+                "block_sparse_moe.experts.7.w2"
+            ]
+        ],
+        "pad_token_id": null,
+        "quant_method": "gptq",
         "sym": true,
+        "tokenizer": null,
         "true_sequential": true,
-        "model_name_or_path": null,
-        "model_file_base_name": "model",
-        "quant_method": "gptq"
-    }
-}
+        "use_cuda_fp16": false,
+        "use_exllama": true
+    },
+    "rms_norm_eps": 1e-05,
+    "rope_theta": 1000000.0,
+    "router_aux_loss_coef": 0.02,
+    "sliding_window": 4096,
+    "tie_word_embeddings": false,
+    "torch_dtype": "bfloat16",
+    "transformers_version": "4.37.0.dev0",
+    "use_cache": true,
+    "vocab_size": 32000
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.37.0.dev0"
+}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f9ef92b3d16e82e3de385999843d9cf49cd3a4e8454d2cac47273a3ce7924d6
-size 23811549344
+oid sha256:574743b06845d2b79545c257ecbe5932e885171ff956338d3cb9bdf49d8f1ae4
+size 23811551008

quantize_config.json

@@ -1,10 +1,8 @@
 {
-  "bits": 4,
-  "group_size": -1,
-  "damp_percent": 0.1,
-  "desc_act": true,
-  "sym": true,
-  "true_sequential": true,
-  "model_name_or_path": null,
-  "model_file_base_name": "model"
+    "bits": 4,
+    "group_size": -1,
+    "damp_percent": 0.1,
+    "desc_act": true,
+    "sym": true,
+    "true_sequential": true
 }