Upload folder using huggingface_hub

chat_template.jinja

@@ -0,0 +1,85 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set content = message.content %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in message.content %}
+                {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
+                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -11,6 +11,48 @@
   "hidden_size": 5120,
   "initializer_range": 0.02,
   "intermediate_size": 17408,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
   "max_position_embeddings": 40960,
   "max_window_layers": 40,
   "model_type": "qwen3",
@@ -20,19 +62,22 @@
   "quantization_config": {
     "bits": 8,
     "checkpoint_format": "gptq",
-    "desc_act": true,
+    "desc_act": false,
     "group_size": 128,
+    "hyb_act": false,
     "lm_head": false,
     "meta": {
-      "damp_auto_increment": 0.0025,
-      "damp_percent": 0.01,
+      "damp_auto_increment": 0.01,
+      "damp_percent": 0.05,
       "mse": 0.0,
       "quantizer": [
-        "gptqmodel:2.2.0"
+        "gptqmodel:4.0.0"
       ],
       "static_groups": false,
       "true_sequential": true,
-      "uri": "https://github.com/modelcloud/gptqmodel"
+      "uri": "https://github.com/modelcloud/gptqmodel",
+      "v2": false,
+      "v2_alpha": 0.25
     },
     "pack_dtype": "int32",
     "quant_method": "gptq",
@@ -44,7 +89,7 @@
   "sliding_window": null,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.51.3",
+  "transformers_version": "4.55.4",
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936

generation_config.json

@@ -9,5 +9,5 @@
   "temperature": 0.6,
   "top_k": 20,
   "top_p": 0.95,
-  "transformers_version": "4.51.3"
+  "transformers_version": "4.55.4"
 }

model-00001-of-00005.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d43cbcd8ed238fe9cccd78c1f61463db407dd876996bcb1a78e0d81cc91e7d69
+oid sha256:bdc23ab6b58f003066fc85b0ef8fdd10f01d8bc954d677b2c8363471b2b24b17
 size 3988118112

model-00002-of-00005.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:73604aa1c7798f8c5e2b5a27c9d420506ab4752c304fa21a4b3009c6b6433e40
+oid sha256:711386c2c2b48fb79bd8b6178255007a0cff64fc92f3c7aef5dbfe223d33d1b5
 size 3994646384

model-00003-of-00005.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55da1d07e76e5c002bf8f39c7a72769d54bdd52ff063dcb8909448590354255a
+oid sha256:5095ba71765947d28d0d2fd5ab7c7a2326656863639bf3969fef314f8fa20d90
 size 3967811360

model-00004-of-00005.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6df050e52713927d570cd4322c2812fd16d32d97bc18d5fb60f64dc30b085c7e
+oid sha256:cc4e23fdc88c11517b662a7393b3010dd438c359c1fd76524d7c0b1d07968e6a
 size 3135659096

quant_log.csv

@@ -0,0 +1,281 @@
+layer,module,loss,samples,damp,time
+0,self_attn.k_proj,0.0000000000,0.05000,1.608
+0,self_attn.v_proj,0.0000000000,0.05000,1.271
+0,self_attn.q_proj,0.0000000001,0.05000,1.268
+0,self_attn.o_proj,0.0000000003,0.05000,1.243
+0,mlp.gate_proj,0.0000000149,0.05000,1.288
+0,mlp.up_proj,0.0000000041,0.05000,1.275
+0,mlp.down_proj,0.0000000021,0.05000,5.248
+1,self_attn.k_proj,0.0000000000,0.05000,1.212
+1,self_attn.v_proj,0.0000000000,0.05000,1.233
+1,self_attn.q_proj,0.0000000002,0.05000,1.236
+1,self_attn.o_proj,0.0000000002,0.05000,1.271
+1,mlp.gate_proj,0.0000002356,0.05000,1.311
+1,mlp.up_proj,0.0000000629,0.05000,1.349
+1,mlp.down_proj,0.0000000041,0.05000,5.250
+2,self_attn.k_proj,0.0000000001,0.05000,1.243
+2,self_attn.v_proj,0.0000000001,0.05000,1.248
+2,self_attn.q_proj,0.0000000004,0.05000,1.293
+2,self_attn.o_proj,0.0000000005,0.05000,1.230
+2,mlp.gate_proj,0.0000003456,0.05000,1.310
+2,mlp.up_proj,0.0000001293,0.05000,1.350
+2,mlp.down_proj,0.0000000034,0.05000,5.178
+3,self_attn.k_proj,0.0000000002,0.05000,1.385
+3,self_attn.v_proj,0.0000000002,0.05000,1.366
+3,self_attn.q_proj,0.0000000007,0.05000,1.383
+3,self_attn.o_proj,0.0000000006,0.05000,1.285
+3,mlp.gate_proj,0.0000004641,0.05000,1.489
+3,mlp.up_proj,0.0000001930,0.05000,1.492
+3,mlp.down_proj,0.0000000064,0.05000,5.132
+4,self_attn.k_proj,0.0000000003,0.05000,1.237
+4,self_attn.v_proj,0.0000000003,0.05000,1.247
+4,self_attn.q_proj,0.0000000012,0.05000,1.250
+4,self_attn.o_proj,0.0000000008,0.05000,1.287
+4,mlp.gate_proj,0.0000006685,0.05000,1.799
+4,mlp.up_proj,0.0000003718,0.05000,1.796
+4,mlp.down_proj,0.0000000080,0.05000,5.095
+5,self_attn.k_proj,0.0000000003,0.05000,1.245
+5,self_attn.v_proj,0.0000000004,0.05000,1.220
+5,self_attn.q_proj,0.0000000014,0.05000,1.239
+5,self_attn.o_proj,0.0000000010,0.05000,1.237
+5,mlp.gate_proj,0.0000007204,0.05000,2.094
+5,mlp.up_proj,0.0000002762,0.05000,2.086
+5,mlp.down_proj,0.0000000159,0.05000,5.226
+6,self_attn.k_proj,0.0000000005,0.05000,1.272
+6,self_attn.v_proj,0.0000000006,0.05000,1.216
+6,self_attn.q_proj,0.0000000023,0.05000,1.229
+6,self_attn.o_proj,0.0000000014,0.05000,1.272
+6,mlp.gate_proj,0.0000010165,0.05000,2.343
+6,mlp.up_proj,0.0000005455,0.05000,2.312
+6,mlp.down_proj,0.0000030633,0.05000,5.788
+7,self_attn.k_proj,0.0000000022,0.05000,1.278
+7,self_attn.v_proj,0.0000000024,0.05000,1.246
+7,self_attn.q_proj,0.0000000095,0.05000,1.363
+7,self_attn.o_proj,0.0000000033,0.05000,1.337
+7,mlp.gate_proj,0.0000009015,0.05000,2.519
+7,mlp.up_proj,0.0000004249,0.05000,2.498
+7,mlp.down_proj,0.0000000286,0.05000,5.549
+8,self_attn.k_proj,0.0000000025,0.05000,1.270
+8,self_attn.v_proj,0.0000000029,0.05000,1.227
+8,self_attn.q_proj,0.0000000109,0.05000,1.263
+8,self_attn.o_proj,0.0000000039,0.05000,1.279
+8,mlp.gate_proj,0.0000002529,0.05000,2.641
+8,mlp.up_proj,0.0000001376,0.05000,2.604
+8,mlp.down_proj,0.0000000270,0.05000,5.703
+9,self_attn.k_proj,0.0000000020,0.05000,1.250
+9,self_attn.v_proj,0.0000000021,0.05000,1.201
+9,self_attn.q_proj,0.0000000083,0.05000,1.245
+9,self_attn.o_proj,0.0000000031,0.05000,1.271
+9,mlp.gate_proj,0.0000001374,0.05000,2.712
+9,mlp.up_proj,0.0000001229,0.05000,2.666
+9,mlp.down_proj,0.0000000360,0.05000,5.699
+10,self_attn.k_proj,0.0000000032,0.05000,1.237
+10,self_attn.v_proj,0.0000000037,0.05000,1.202
+10,self_attn.q_proj,0.0000000146,0.05000,1.222
+10,self_attn.o_proj,0.0000000055,0.05000,1.265
+10,mlp.gate_proj,0.0000001438,0.05000,2.729
+10,mlp.up_proj,0.0000001321,0.05000,2.691
+10,mlp.down_proj,0.0000000309,0.05000,5.662
+11,self_attn.k_proj,0.0000000047,0.05000,1.236
+11,self_attn.v_proj,0.0000000054,0.05000,1.206
+11,self_attn.q_proj,0.0000000213,0.05000,1.242
+11,self_attn.o_proj,0.0000000080,0.05000,1.279
+11,mlp.gate_proj,0.0000001702,0.05000,2.725
+11,mlp.up_proj,0.0000001570,0.05000,2.679
+11,mlp.down_proj,0.0000000382,0.05000,5.663
+12,self_attn.k_proj,0.0000000038,0.05000,1.270
+12,self_attn.v_proj,0.0000000043,0.05000,1.221
+12,self_attn.q_proj,0.0000000168,0.05000,1.246
+12,self_attn.o_proj,0.0000000083,0.05000,1.287
+12,mlp.gate_proj,0.0000001990,0.05000,2.718
+12,mlp.up_proj,0.0000001818,0.05000,2.679
+12,mlp.down_proj,0.0000000475,0.05000,5.667
+13,self_attn.k_proj,0.0000000083,0.05000,1.218
+13,self_attn.v_proj,0.0000000095,0.05000,1.204
+13,self_attn.q_proj,0.0000000373,0.05000,1.239
+13,self_attn.o_proj,0.0000000132,0.05000,1.437
+13,mlp.gate_proj,0.0000002398,0.05000,2.726
+13,mlp.up_proj,0.0000002055,0.05000,2.682
+13,mlp.down_proj,0.0000000539,0.05000,5.732
+14,self_attn.k_proj,0.0000000060,0.05000,1.249
+14,self_attn.v_proj,0.0000000070,0.05000,1.207
+14,self_attn.q_proj,0.0000000262,0.05000,1.231
+14,self_attn.o_proj,0.0000000124,0.05000,1.283
+14,mlp.gate_proj,0.0000002277,0.05000,2.766
+14,mlp.up_proj,0.0000002028,0.05000,2.716
+14,mlp.down_proj,0.0000000555,0.05000,5.775
+15,self_attn.k_proj,0.0000000054,0.05000,1.234
+15,self_attn.v_proj,0.0000000057,0.05000,1.239
+15,self_attn.q_proj,0.0000000232,0.05000,1.236
+15,self_attn.o_proj,0.0000000120,0.05000,1.333
+15,mlp.gate_proj,0.0000002143,0.05000,2.781
+15,mlp.up_proj,0.0000002041,0.05000,2.748
+15,mlp.down_proj,0.0000000569,0.05000,5.749
+16,self_attn.k_proj,0.0000000069,0.05000,1.247
+16,self_attn.v_proj,0.0000000078,0.05000,1.218
+16,self_attn.q_proj,0.0000000293,0.05000,1.238
+16,self_attn.o_proj,0.0000000160,0.05000,1.282
+16,mlp.gate_proj,0.0000001931,0.05000,2.763
+16,mlp.up_proj,0.0000001971,0.05000,2.714
+16,mlp.down_proj,0.0000000595,0.05000,5.776
+17,self_attn.k_proj,0.0000000070,0.05000,1.259
+17,self_attn.v_proj,0.0000000076,0.05000,1.244
+17,self_attn.q_proj,0.0000000310,0.05000,1.272
+17,self_attn.o_proj,0.0000000174,0.05000,1.294
+17,mlp.gate_proj,0.0000002041,0.05000,2.790
+17,mlp.up_proj,0.0000002107,0.05000,2.739
+17,mlp.down_proj,0.0000000591,0.05000,5.725
+18,self_attn.k_proj,0.0000000099,0.05000,1.236
+18,self_attn.v_proj,0.0000000110,0.05000,1.231
+18,self_attn.q_proj,0.0000000449,0.05000,1.244
+18,self_attn.o_proj,0.0000000171,0.05000,1.300
+18,mlp.gate_proj,0.0000002128,0.05000,2.755
+18,mlp.up_proj,0.0000002273,0.05000,2.717
+18,mlp.down_proj,0.0000000713,0.05000,5.824
+19,self_attn.k_proj,0.0000000133,0.05000,1.419
+19,self_attn.v_proj,0.0000000144,0.05000,1.370
+19,self_attn.q_proj,0.0000000573,0.05000,1.400
+19,self_attn.o_proj,0.0000000215,0.05000,3.100
+19,mlp.gate_proj,0.0000002382,0.05000,2.840
+19,mlp.up_proj,0.0000002524,0.05000,2.880
+19,mlp.down_proj,0.0000001835,0.05000,11.147
+20,self_attn.k_proj,0.0000000244,0.05000,2.686
+20,self_attn.v_proj,0.0000000304,0.05000,2.694
+20,self_attn.q_proj,0.0000001156,0.05000,2.743
+20,self_attn.o_proj,0.0000000301,0.05000,2.666
+20,mlp.gate_proj,0.0000002484,0.05000,3.245
+20,mlp.up_proj,0.0000002656,0.05000,3.233
+20,mlp.down_proj,0.0000001017,0.05000,12.341
+21,self_attn.k_proj,0.0000000251,0.05000,2.986
+21,self_attn.v_proj,0.0000000264,0.05000,2.962
+21,self_attn.q_proj,0.0000001019,0.05000,3.059
+21,self_attn.o_proj,0.0000000396,0.05000,2.976
+21,mlp.gate_proj,0.0000002622,0.05000,3.496
+21,mlp.up_proj,0.0000002757,0.05000,3.500
+21,mlp.down_proj,0.0000001119,0.05000,11.943
+22,self_attn.k_proj,0.0000000238,0.05000,2.946
+22,self_attn.v_proj,0.0000000296,0.05000,2.853
+22,self_attn.q_proj,0.0000001088,0.05000,2.927
+22,self_attn.o_proj,0.0000000381,0.05000,2.933
+22,mlp.gate_proj,0.0000002831,0.05000,3.603
+22,mlp.up_proj,0.0000003007,0.05000,3.516
+22,mlp.down_proj,0.0000001143,0.05000,12.494
+23,self_attn.k_proj,0.0000000342,0.05000,2.872
+23,self_attn.v_proj,0.0000000455,0.05000,2.833
+23,self_attn.q_proj,0.0000001655,0.05000,2.872
+23,self_attn.o_proj,0.0000000303,0.05000,3.006
+23,mlp.gate_proj,0.0000002891,0.05000,3.699
+23,mlp.up_proj,0.0000003093,0.05000,3.486
+23,mlp.down_proj,0.0000001386,0.05000,12.495
+24,self_attn.k_proj,0.0000000418,0.05000,2.969
+24,self_attn.v_proj,0.0000000572,0.05000,3.049
+24,self_attn.q_proj,0.0000002142,0.05000,3.041
+24,self_attn.o_proj,0.0000000520,0.05000,3.151
+24,mlp.gate_proj,0.0000003189,0.05000,3.735
+24,mlp.up_proj,0.0000003336,0.05000,3.711
+24,mlp.down_proj,0.0000001439,0.05000,12.620
+25,self_attn.k_proj,0.0000000434,0.05000,1.259
+25,self_attn.v_proj,0.0000000548,0.05000,1.228
+25,self_attn.q_proj,0.0000002045,0.05000,1.253
+25,self_attn.o_proj,0.0000000486,0.05000,1.285
+25,mlp.gate_proj,0.0000003654,0.05000,3.381
+25,mlp.up_proj,0.0000003748,0.05000,3.274
+25,mlp.down_proj,0.0000001957,0.05000,6.119
+26,self_attn.k_proj,0.0000000447,0.05000,1.252
+26,self_attn.v_proj,0.0000000600,0.05000,1.214
+26,self_attn.q_proj,0.0000002259,0.05000,1.230
+26,self_attn.o_proj,0.0000000501,0.05000,1.267
+26,mlp.gate_proj,0.0000004381,0.05000,2.894
+26,mlp.up_proj,0.0000004364,0.05000,2.844
+26,mlp.down_proj,0.0000003177,0.05000,5.673
+27,self_attn.k_proj,0.0000000808,0.05000,1.259
+27,self_attn.v_proj,0.0000001062,0.05000,1.214
+27,self_attn.q_proj,0.0000003821,0.05000,1.235
+27,self_attn.o_proj,0.0000000942,0.05000,1.285
+27,mlp.gate_proj,0.0000005449,0.05000,2.707
+27,mlp.up_proj,0.0000005444,0.05000,2.679
+27,mlp.down_proj,0.0000004879,0.05000,5.670
+28,self_attn.k_proj,0.0000001069,0.05000,1.244
+28,self_attn.v_proj,0.0000001591,0.05000,1.218
+28,self_attn.q_proj,0.0000005816,0.05000,1.240
+28,self_attn.o_proj,0.0000001124,0.05000,1.282
+28,mlp.gate_proj,0.0000006669,0.05000,2.688
+28,mlp.up_proj,0.0000006805,0.05000,2.661
+28,mlp.down_proj,0.0000006141,0.05000,5.578
+29,self_attn.k_proj,0.0000001652,0.05000,1.224
+29,self_attn.v_proj,0.0000002137,0.05000,1.218
+29,self_attn.q_proj,0.0000007917,0.05000,1.223
+29,self_attn.o_proj,0.0000000944,0.05000,1.286
+29,mlp.gate_proj,0.0000007430,0.05000,2.691
+29,mlp.up_proj,0.0000007675,0.05000,2.657
+29,mlp.down_proj,0.0000009056,0.05000,5.653
+30,self_attn.k_proj,0.0000002540,0.05000,1.300
+30,self_attn.v_proj,0.0000003661,0.05000,1.221
+30,self_attn.q_proj,0.0000012225,0.05000,1.254
+30,self_attn.o_proj,0.0000002047,0.05000,1.293
+30,mlp.gate_proj,0.0000009594,0.05000,2.721
+30,mlp.up_proj,0.0000009889,0.05000,2.666
+30,mlp.down_proj,0.0000011860,0.05000,5.710
+31,self_attn.k_proj,0.0000002662,0.05000,1.266
+31,self_attn.v_proj,0.0000003309,0.05000,1.229
+31,self_attn.q_proj,0.0000012580,0.05000,1.253
+31,self_attn.o_proj,0.0000001825,0.05000,1.279
+31,mlp.gate_proj,0.0000010435,0.05000,2.735
+31,mlp.up_proj,0.0000010983,0.05000,2.693
+31,mlp.down_proj,0.0000014923,0.05000,5.708
+32,self_attn.k_proj,0.0000004588,0.05000,1.257
+32,self_attn.v_proj,0.0000006749,0.05000,1.223
+32,self_attn.q_proj,0.0000022000,0.05000,1.256
+32,self_attn.o_proj,0.0000001838,0.05000,1.291
+32,mlp.gate_proj,0.0000011835,0.05000,2.746
+32,mlp.up_proj,0.0000012659,0.05000,2.906
+32,mlp.down_proj,0.0000019001,0.05000,5.747
+33,self_attn.k_proj,0.0000005869,0.05000,1.227
+33,self_attn.v_proj,0.0000008845,0.05000,1.226
+33,self_attn.q_proj,0.0000029721,0.05000,1.245
+33,self_attn.o_proj,0.0000002444,0.05000,1.273
+33,mlp.gate_proj,0.0000012551,0.05000,2.778
+33,mlp.up_proj,0.0000013810,0.05000,2.740
+33,mlp.down_proj,0.0000024191,0.05000,5.714
+34,self_attn.k_proj,0.0000009436,0.05000,1.226
+34,self_attn.v_proj,0.0000014251,0.05000,1.244
+34,self_attn.q_proj,0.0000047080,0.05000,1.284
+34,self_attn.o_proj,0.0000002777,0.05000,1.266
+34,mlp.gate_proj,0.0000014021,0.05000,2.765
+34,mlp.up_proj,0.0000015736,0.05000,2.723
+34,mlp.down_proj,0.0000030276,0.05000,5.794
+35,self_attn.k_proj,0.0000012035,0.05000,1.236
+35,self_attn.v_proj,0.0000019715,0.05000,1.209
+35,self_attn.q_proj,0.0000061430,0.05000,1.227
+35,self_attn.o_proj,0.0000002815,0.05000,1.280
+35,mlp.gate_proj,0.0000015192,0.05000,2.764
+35,mlp.up_proj,0.0000017237,0.05000,2.725
+35,mlp.down_proj,0.0000044798,0.05000,6.006
+36,self_attn.k_proj,0.0000012551,0.05000,1.238
+36,self_attn.v_proj,0.0000019508,0.05000,1.226
+36,self_attn.q_proj,0.0000057919,0.05000,1.238
+36,self_attn.o_proj,0.0000007864,0.05000,1.282
+36,mlp.gate_proj,0.0000016147,0.05000,2.746
+36,mlp.up_proj,0.0000018898,0.05000,2.715
+36,mlp.down_proj,0.0000064629,0.05000,5.655
+37,self_attn.k_proj,0.0000016820,0.05000,1.387
+37,self_attn.v_proj,0.0000029921,0.05000,1.228
+37,self_attn.q_proj,0.0000083882,0.05000,1.269
+37,self_attn.o_proj,0.0000008105,0.05000,1.375
+37,mlp.gate_proj,0.0000017379,0.05000,2.764
+37,mlp.up_proj,0.0000020626,0.05000,2.722
+37,mlp.down_proj,0.0000090136,0.05000,5.689
+38,self_attn.k_proj,0.0000017496,0.05000,1.223
+38,self_attn.v_proj,0.0000032497,0.05000,1.208
+38,self_attn.q_proj,0.0000083224,0.05000,1.225
+38,self_attn.o_proj,0.0000017614,0.05000,1.280
+38,mlp.gate_proj,0.0000022542,0.05000,2.758
+38,mlp.up_proj,0.0000025480,0.05000,2.713
+38,mlp.down_proj,0.0000141993,0.05000,5.750
+39,self_attn.k_proj,0.0000008697,0.05000,1.233
+39,self_attn.v_proj,0.0000013233,0.05000,1.214
+39,self_attn.q_proj,0.0000040708,0.05000,1.232
+39,self_attn.o_proj,0.0000014657,0.05000,1.255
+39,mlp.gate_proj,0.0000032426,0.05000,2.746
+39,mlp.up_proj,0.0000035667,0.05000,2.682
+39,mlp.down_proj,0.0000380214,0.05000,5.708

quantize_config.json

@@ -1,7 +1,8 @@
 {
   "bits": 8,
   "group_size": 128,
-  "desc_act": true,
+  "desc_act": false,
+  "hyb_act": false,
   "sym": true,
   "lm_head": false,
   "quant_method": "gptq",
@@ -9,13 +10,15 @@
   "pack_dtype": "int32",
   "meta": {
     "quantizer": [
-      "gptqmodel:2.2.0"
+      "gptqmodel:4.0.0"
     ],
     "uri": "https://github.com/modelcloud/gptqmodel",
-    "damp_percent": 0.01,
-    "damp_auto_increment": 0.0025,
+    "damp_percent": 0.05,
+    "damp_auto_increment": 0.01,
     "static_groups": false,
     "true_sequential": true,
-    "mse": 0.0
+    "mse": 0.0,
+    "v2": false,
+    "v2_alpha": 0.25
   }
 }

tokenizer_config.json

@@ -227,7 +227,6 @@
     "<|video_pad|>"
   ],
   "bos_token": null,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in message.content %}\n                {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<|im_end|>",
   "errors": "replace",