Upload folder using huggingface_hub

.ms_upload_cache

@@ -0,0 +1 @@
+{"version": 3, "repo_id": "SLAIAITP/DeepSeek-V4-Pro-OR", "files": {"chat_template.jinja|1782975878.753825|2118": {"hash": "19f422f71b17f857f71b031469f92640198f5f3e8b42aeb035b14fb919974731", "size": 2118, "status": "c"}, "generation_config.json|1777019529.1717968|170": {"hash": "5fccff80f55a4d455bbe516bdd552edf3e9623df95e99fbf2a3c3389fdf91af0", "size": 170, "status": "c"}, "config.json|1783929216.7320757|1886": {"hash": "de73d54a88903376b4001be2055967d6eb1c2f2eef0a03275e0d7a3dcf066ccc", "size": 1886, "status": "c"}, "configuration.json|1777872707.049769|48": {"hash": "a78221ae93dd8977b26fcd1d106fa5ad7e7f1fd3a7d93bfd9d6fba5ed00ec4d2", "size": 48, "status": "c"}, ".ipynb_checkpoints/upload-checkpoint.py|1784877124.9296815|1732": {"hash": "0a952f9bdb0dfac24ac86a72d77978901d400c425185a32042519f0a4db1ae55", "size": 1732, "status": "c"}, "config_0712.json|1783855574.2283494|1886": {"hash": "de73d54a88903376b4001be2055967d6eb1c2f2eef0a03275e0d7a3dcf066ccc", "size": 1886, "status": "c"}, "config_org.json|1779098365.718372|2079": {"hash": "d9b8282dc513c501c666fe4039622ddecbc0ade828f99badd91adc2897fe8812", "size": 2079, "status": "c"}, "model-00002-of-000062.safetensors|1784142939.6177006|51495649788": {"hash": "9bfa78c6234054305a36d06122d3b80db87532981ae5573a9f57039199ebc757", "size": 51495649788, "status": "c"}, "model-00005-of-000062.safetensors|1784142900.5097308|51537429380": {"hash": "d576955ab885a6f82708e706c139f59238ed207de775ba772beb2d5a6f861b04", "size": 51537429380, "status": "c"}, "model-00006-of-000062.safetensors|1784143024.2639053|51489445852": {"hash": "9e49f481c6bd3c9d1e90d271fbd01f2b9a15b59b4905105b743ad4a8caca7e7b", "size": 51489445852, "status": "c"}, "model-00007-of-000062.safetensors|1784143136.322542|51537429380": {"hash": "bde8975b0707479555da28170b1dbe5a728cb8dfbc34902af3a855ec4388b0ef", "size": 51537429380, "status": "c"}, "model-00008-of-000062.safetensors|1784147501.7918267|51489445852": {"hash": "e5307f721d2658005fa7bf23070a5c04d8d0f26f590c918ea6e51a86f195d849", "size": 51489445852, "status": "c"}, "model-00004-of-000062.safetensors|1784142781.3327672|51489445852": {"hash": "37f9255b571cc3b531b29e0b9bea52a622ab7228bc2133700fd0b9b5bde7cdc7", "size": 51489445852, "status": "c"}, "model-00003-of-000062.safetensors|1784143104.117064|51543633316": {"hash": "57d2a76d68384c481987236a71b1936777605545db5fddf5141211d1d3960c36", "size": 51543633316, "status": "c"}, "model-00001-of-000062.safetensors|1784142796.9312716|53349008108": {"hash": "7c8a4cd1fdf0253dbd2d85a8eb4bc1758e7772cde8ff583bbbb8be0c76df195c", "size": 53349008108, "status": "c"}, "model-00009-of-000062.safetensors|1784147647.826568|51537429380": {"hash": "92a9b76991fe2fe10ac8a1a1a5bd36437e6aa8d6a69abcd84e26d0c92880f020", "size": 51537429380, "status": "c"}, "model-00010-of-000062.safetensors|1784147838.5781012|51489445852": {"hash": "b487ef65e5de403196d808ab9a2c4a8f09573974bf1f9bce7e0e5c40d750b89e", "size": 51489445852, "status": "c"}, "model-00011-of-000062.safetensors|1784148025.7857368|51537430564": {"hash": "7bc08b3150750a0494ef15a43c3e0259b75c3f9f2d6cfcf8d868df4266a7d34a", "size": 51537430564, "status": "c"}, "model-00012-of-000062.safetensors|1784147597.602396|51489447028": {"hash": "fcce73b85a23d15c327de7a4552e94db84d8def0e7520117c9b9880331443bd4", "size": 51489447028, "status": "c"}, "model-00013-of-000062.safetensors|1784147780.0649176|51537430564": {"hash": "32bcb55c099db4915468c76ab2ed4490c9d75d6abfbb8a8bb757df7d46dc05d6", "size": 51537430564, "status": "c"}, "model-00014-of-000062.safetensors|1784147939.068244|51489447028": {"hash": "e5aba4f4ed60573497a6e169b762395a48b56a1be6653b122cac0d27ae09db2a", "size": 51489447028, "status": "c"}, "model-00015-of-000062.safetensors|1784148103.764836|51537430564": {"hash": "6fcf51e978f29bb05b3df25e8bc5a7e3788302ebd6ed364396fe139b50835f47", "size": 51537430564, "status": "c"}, "model-00016-of-000062.safetensors|1784152104.097912|51489447028": {"hash": "077013dea389011cef57177fe417d0fdcfbf26f621ce5264b1b49694845b3812", "size": 51489447028, "status": "c"}, "model-00017-of-000062.safetensors|1784152292.4470525|51537430564": {"hash": "fe404fb0dd1a3bbfc94cc4f356bcdf09e97f6af123fed497e663bacf00a489fc", "size": 51537430564, "status": "c"}, "model-00018-of-000062.safetensors|1784152462.862392|51489447028": {"hash": "20d7811fb00bb9e5ac3863e721b608cad681ba8b490ccfc229f5ebe1a498f16a", "size": 51489447028, "status": "c"}, "model-00019-of-000062.safetensors|1784152661.8143632|51537430564": {"hash": "f0228eb5d1e4107f8b03502efd4495bf842b57ee42738dce35a2599c4d134312", "size": 51537430564, "status": "c"}, "model-00020-of-000062.safetensors|1784152322.1304011|51489447028": {"hash": "2b32d91754e5dce2a18f91a86d9d0add1550ba5cf5cf3ad0c9246fb8600910c1", "size": 51489447028, "status": "c"}, "model-00021-of-000062.safetensors|1784152502.5606036|51537430564": {"hash": "0d93e2725c723e66673144fb84e01b6f947ebea96ca7e430ff8ebd8a7e794551", "size": 51537430564, "status": "c"}, "model-00022-of-000062.safetensors|1784152680.113828|51489447028": {"hash": "4584aafeb8ce3185cf56ac7d7d8ec95c854a26da7e8fe4bdb17990343fd8b1de", "size": 51489447028, "status": "c"}, "model-00023-of-000062.safetensors|1784152826.6594865|51537430564": {"hash": "85a02a7658907a17fd14aaede44adc9239a5cb9e297b6dff91f96a12dcf11805", "size": 51537430564, "status": "c"}, "model-00024-of-000062.safetensors|1784157578.1422243|51489447028": {"hash": "b5fb5d5054264f9a73ee8dd2e552aa03166d1b86e072fc683794ebda0a635ba6", "size": 51489447028, "status": "c"}, "model-00025-of-000062.safetensors|1784157767.0702994|51537430564": {"hash": "0cdadf3a8cc707fc0313e7667c0e94c5242479b3d291c89cbcbfed6400ad3438", "size": 51537430564, "status": "c"}, "model-00026-of-000062.safetensors|1784157944.7719932|51489447028": {"hash": "4b415282c6921d47177e9893f4a0361e908fe8ce94c5cee18cb9ce9b747577fb", "size": 51489447028, "status": "c"}, "model-00027-of-000062.safetensors|1784158109.773555|51537430564": {"hash": "61b8a2ef58bfab9c9b7ba51d42f698ffe2d2dfd4aae37d53acacbfc67d656719", "size": 51537430564, "status": "c"}, "model-00028-of-000062.safetensors|1784157720.48404|51489447028": {"hash": "3c08db5eafe23a134b88163ee64dc902cd62e7289ef80d3df5afa96244cf14f2", "size": 51489447028, "status": "c"}, "model-00029-of-000062.safetensors|1784157891.4676683|51537430564": {"hash": "7fac05d41c3db396c1a255b8422573cbec9b26085d024d9fc3d467c9f9450203", "size": 51537430564, "status": "c"}, "model-00030-of-000062.safetensors|1784158038.2214131|51489447028": {"hash": "e23e3526f062f455964f82591f4e2c22aa7045cc724fd68a640f5d58c8ec2bcf", "size": 51489447028, "status": "c"}, "model-00031-of-000062.safetensors|1784158174.1049225|51537430564": {"hash": "45c131ec0a359467fa47c32241badb26283c5acd9e06ffaf35361dc93876209d", "size": 51537430564, "status": "c"}, "model-00032-of-000062.safetensors|1784144855.1125023|51489447028": {"hash": "913cf75c0b68b3f897787390244d77dd0b35a3c5a4f7900b227d629f420f5bdd", "size": 51489447028, "status": "c"}, "model-00033-of-000062.safetensors|1784145029.7772717|51537430564": {"hash": "752213f4497b3f573aab11023edd7a76c2421e32e2f7b5ba092aca9475173b2d", "size": 51537430564, "status": "c"}, "model-00034-of-000062.safetensors|1784145181.8614087|51489447028": {"hash": "26c4cbe622de47f67c4df861b8d0c2fe753939e3ce51acd2ddf38c78aa517bc9", "size": 51489447028, "status": "c"}, "model-00035-of-000062.safetensors|1784145330.9466772|51537430564": {"hash": "b25569094ef4e43c149fb137e9ce1b356de447322389d3da7b9f359c613938c2", "size": 51537430564, "status": "c"}, "model-00036-of-000062.safetensors|1784144955.1158626|51489447028": {"hash": "4f3fbc66d3474e0e4083a239463c019cc55fd39d13c22e2a6959e5fc59195487", "size": 51489447028, "status": "c"}, "model-00037-of-000062.safetensors|1784145124.9503434|51537430564": {"hash": "4f0baabbf814036cc424fde6e25f78de4345b323bb7de002439da73ae9a4de11", "size": 51537430564, "status": "c"}, "model-00038-of-000062.safetensors|1784145279.2123113|51489447028": {"hash": "314b16b19db0715907829e46207607f6b62411b8d599e870e42682ed3bc44e8a", "size": 51489447028, "status": "c"}, "model-00039-of-000062.safetensors|1784145445.7135775|51537430564": {"hash": "d4874f6cb26f181505fa8590e409c3db8924fa19ee0fd521c76046d180d6a1ea", "size": 51537430564, "status": "c"}, "model-00040-of-000062.safetensors|1784149333.470021|51489447028": {"hash": "74d9f898ed1afe33390f0f6f195797c5dc6c1b6e62554854dfd898210e663353", "size": 51489447028, "status": "c"}, "model-00041-of-000062.safetensors|1784149522.199945|51537430564": {"hash": "33ef0a0c800f029065592a03bd15a68d1eea9b4096c00ed16ba821d780eadc9b", "size": 51537430564, "status": "c"}, "model-00042-of-000062.safetensors|1784149716.342556|51489447028": {"hash": "25376fbf06ae58d521012486ebed9e08fd3eccd718e7c6095ed6b89bf845c808", "size": 51489447028, "status": "c"}, "model-00043-of-000062.safetensors|1784149888.4503586|51537430564": {"hash": "8911cee8463c7780c576bcfab1d18d57fc3899b269d2d8304cdc98f5f5bbff74", "size": 51537430564, "status": "c"}, "model-00044-of-000062.safetensors|1784149738.6376266|51489447028": {"hash": "080a476122f1d697d10518acf8434c3b99ad16bf8ab30f80febd76f8776a64fe", "size": 51489447028, "status": "c"}, "model-00045-of-000062.safetensors|1784149885.6284966|51537430564": {"hash": "85f87600c93054cbe7896150952da4db2ae0c85fac9610b3c6ed6b8ae3028d23", "size": 51537430564, "status": "c"}, "model-00046-of-000062.safetensors|1784149983.5352805|51489447028": {"hash": "a14019a7f932cf867cd3314f946f936b490c7f940bacc2b1237484a283820ff6", "size": 51489447028, "status": "c"}, "model-00047-of-000062.safetensors|1784150116.492498|51537430564": {"hash": "fa1a6a854b2fe1a66f93ac0d17c78b3a2c2788f4529509c3ee68373f3616706e", "size": 51537430564, "status": "c"}, "model-00048-of-000062.safetensors|1784154803.5291865|51489447028": {"hash": "cc2eaea59443ec93d321219055eb165e70bf51764f0889cf905a922dbce56e9f", "size": 51489447028, "status": "c"}, "model-00049-of-000062.safetensors|1784155018.4155827|51537430564": {"hash": "ce57a706db9f3a42f01898eeb5aaa7827af0e06b9ba008f25d204fc39cc0dad7", "size": 51537430564, "status": "c"}, "model-00050-of-000062.safetensors|1784155187.806661|51489447028": {"hash": "0ea693ef7432f1384224002b44bcaae503349f136279230a64a3d5f9ea3d534f", "size": 51489447028, "status": "c"}, "model-00051-of-000062.safetensors|1784155355.3136983|51537430564": {"hash": "6d5a594b2abcf1ce6a52618526f8770967b6d73cbdb605a2ce7b62609c08b807", "size": 51537430564, "status": "c"}, "model-00052-of-000062.safetensors|1784155008.9776669|51489447028": {"hash": "7b7676bcc67b3e4d6d9c604c8793d7efdd2f62bebc9322227e3b629fa7861eb1", "size": 51489447028, "status": "c"}, "model-00053-of-000062.safetensors|1784155167.8618796|51537430564": {"hash": "a3ed4d5177727068ebbb3b83235b55a20747d2546c1bee14b0fcf71f073cdeeb", "size": 51537430564, "status": "c"}, "model-00054-of-000062.safetensors|1784155301.2734776|51489447028": {"hash": "40dce8f9da388a43655ab7cb1ca3b594dba638603b1f4e2d3fd8530f1afc23ae", "size": 51489447028, "status": "c"}, "model-00055-of-000062.safetensors|1784155440.7349532|51537430564": {"hash": "8c30bea36645a40a12babe55211c416f02da992d4baf347c9c4f4c7229aa7b72", "size": 51537430564, "status": "c"}, "model-00056-of-000062.safetensors|1784160205.5194955|51489447028": {"hash": "9c1646588f7c8cf98e572109afd97ba1920e08ea08178559f14a4e7d7e45098a", "size": 51489447028, "status": "c"}, "model-00057-of-000062.safetensors|1784160399.0733871|51537430564": {"hash": "30ce88bb7de659acc5001c63c54db5cd3c3c3b059b46a05b88acbde4f201fe0e", "size": 51537430564, "status": "c"}, "model-00058-of-000062.safetensors|1784160577.1172574|51489447028": {"hash": "f74390e1a5f9fd78ea91cb8ee3a6dab998679eec4a4e093e9a48bf19a1f218b4", "size": 51489447028, "status": "c"}, "model-00059-of-000062.safetensors|1784160773.288997|51537430564": {"hash": "16869bd8552d3e1a623697d6e1ae43a1bb6374f3ad6dd2ae6ffeb6c316d48134", "size": 51537430564, "status": "c"}, "model-00060-of-000062.safetensors|1784160432.0814743|51489447028": {"hash": "a82212b512a53ecc1fbf0c15ee08e963f842a783ccdb438f739e64c3d419e107", "size": 51489447028, "status": "c"}, "model.safetensors.index.json|1784171486.3493745|6098309": {"hash": "6772c7b0007f6ba81db8f4f3435a65f0c35c27a310a8718bba07d1a7b677b6fb", "size": 6098309, "status": "c"}, "tokenizer.json|1777730073.108539|6367146": {"hash": "8f9f37ca37fdc4f5fd36d5cf4d3b0e8392edb4e894fd10cc0d70b4957c8633cf", "size": 6367146, "status": "c"}, "tokenizer_config.json|1777730073.7466133|801": {"hash": "6ac8c8dc065ed118161d02dd532749ae3f52c243deac27872134fae2f50d8547", "size": 801, "status": "c"}, "model-00061-of-000062.safetensors|1784160621.9456851|53391032958": {"hash": "ae16cc536380a7e685d73d28be867d14b874b2519071035648ebc2ee44e6b206", "size": 53391032958, "status": "c"}, "model-00062-of-000062.safetensors|1784160752.871929|51680423790": {"hash": "abcd3bd2bda7a14a11e2868bf88bcffa188500c1c33ce8541eb84294b2a2494b", "size": 51680423790, "status": "c"}}}

chat_template.jinja

@@ -0,0 +1,61 @@
+{%- if not add_generation_prompt is defined -%}
+  {%- set add_generation_prompt = false -%}
+{%- endif -%}
+{%- if not thinking is defined -%}
+  {%- if enable_thinking is defined -%}
+    {%- set thinking = enable_thinking -%}
+  {%- else -%}
+    {%- set thinking = false -%}
+  {%- endif -%}
+{%- endif -%}
+{%- set ns = namespace(system_prompt='', is_first_sp=true, is_last_user=false, is_only_sys=false, is_prefix=false) -%}
+{%- for message in messages -%}
+  {%- if message['role'] == 'system' -%}
+    {%- if ns.is_first_sp -%}
+      {%- set ns.system_prompt = ns.system_prompt + message['content'] -%}
+      {%- set ns.is_first_sp = false -%}
+    {%- else -%}
+      {%- set ns.system_prompt = ns.system_prompt + '\n\n' + message['content'] -%}
+    {%- endif -%}
+    {%- set ns.is_only_sys = true -%}
+  {%- endif -%}
+{%- endfor -%}
+{{- bos_token -}}{{- ns.system_prompt -}}
+{%- for message in messages -%}
+  {%- if message['role'] == 'user' -%}
+    {%- set ns.is_last_user = true -%}
+    {{- '<｜User｜>' + message['content'] -}}
+  {%- endif -%}
+  {%- if message['role'] == 'assistant' -%}
+    {%- if ns.is_last_user -%}
+      {{- '<｜Assistant｜>' -}}
+      {%- if message['prefix'] is defined and message['prefix'] and thinking -%}
+        {{- '<think>' -}}
+      {%- else -%}
+        {{- '</think>' -}}
+      {%- endif -%}
+    {%- endif -%}
+    {%- if message['prefix'] is defined and message['prefix'] -%}
+      {%- set ns.is_prefix = true -%}
+    {%- endif -%}
+    {%- set ns.is_last_user = false -%}
+    {%- set content = message['content'] -%}
+    {%- if '</think>' in content -%}
+      {%- set content = content.split('</think>', 1)[1] -%}
+    {%- endif -%}
+    {{- content + '<｜end▁of▁sentence｜>' -}}
+  {%- endif -%}
+  {%- if message['role'] != 'system' -%}
+    {%- set ns.is_only_sys = false -%}
+  {%- endif -%}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+  {%- if ns.is_last_user or ns.is_only_sys or not ns.is_prefix -%}
+    {{- '<｜Assistant｜>' -}}
+    {%- if not thinking -%}
+      {{- '</think>' -}}
+    {%- else -%}
+      {{- '<think>' -}}
+    {%- endif -%}
+  {%- endif -%}
+{%- endif -%}

config.json

@@ -0,0 +1,120 @@
+{
+  "architectures": [
+    "DeepseekV4ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "expert_dtype": "fp4",
+  "hc_eps": 1e-06,
+  "hc_mult": 4,
+  "hc_sinkhorn_iters": 20,
+  "head_dim": 512,
+  "hidden_act": "silu",
+  "hidden_size": 7168,
+  "index_head_dim": 128,
+  "index_n_heads": 64,
+  "index_topk": 1024,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 1048576,
+  "model_type": "deepseek_v4",
+  "moe_intermediate_size": 3072,
+  "n_routed_experts": 384,
+  "n_shared_experts": 1,
+  "norm_topk_prob": true,
+  "num_attention_heads": 128,
+  "num_experts_per_tok": 6,
+  "num_hidden_layers": 61,
+  "num_hash_layers": 3,
+  "num_key_value_heads": 1,
+  "num_nextn_predict_layers": 1,
+  "o_groups": 16,
+  "o_lora_rank": 1024,
+  "q_lora_rank": 1536,
+  "qk_rope_head_dim": 64,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "beta_fast": 32,
+    "beta_slow": 1,
+    "factor": 16,
+    "original_max_position_embeddings": 65536,
+    "type": "yarn"
+  },
+  "rope_theta": 10000,
+  "routed_scaling_factor": 2.5,
+  "scoring_func": "sqrtsoftplus",
+  "sliding_window": 128,
+  "swiglu_limit": 10.0,
+  "tie_word_embeddings": false,
+  "topk_method": "noaux_tc",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "vocab_size": 129280,
+  "compress_rope_theta": 160000,
+  "compress_ratios": [
+    128,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    0
+  ]
+}

config_0712.json

@@ -0,0 +1,120 @@
+{
+  "architectures": [
+    "DeepseekV4ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "expert_dtype": "fp4",
+  "hc_eps": 1e-06,
+  "hc_mult": 4,
+  "hc_sinkhorn_iters": 20,
+  "head_dim": 512,
+  "hidden_act": "silu",
+  "hidden_size": 7168,
+  "index_head_dim": 128,
+  "index_n_heads": 64,
+  "index_topk": 1024,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 1048576,
+  "model_type": "deepseek_v4",
+  "moe_intermediate_size": 3072,
+  "n_routed_experts": 384,
+  "n_shared_experts": 1,
+  "norm_topk_prob": true,
+  "num_attention_heads": 128,
+  "num_experts_per_tok": 6,
+  "num_hidden_layers": 61,
+  "num_hash_layers": 3,
+  "num_key_value_heads": 1,
+  "num_nextn_predict_layers": 1,
+  "o_groups": 16,
+  "o_lora_rank": 1024,
+  "q_lora_rank": 1536,
+  "qk_rope_head_dim": 64,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "beta_fast": 32,
+    "beta_slow": 1,
+    "factor": 16,
+    "original_max_position_embeddings": 65536,
+    "type": "yarn"
+  },
+  "rope_theta": 10000,
+  "routed_scaling_factor": 2.5,
+  "scoring_func": "sqrtsoftplus",
+  "sliding_window": 128,
+  "swiglu_limit": 10.0,
+  "tie_word_embeddings": false,
+  "topk_method": "noaux_tc",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "vocab_size": 129280,
+  "compress_rope_theta": 160000,
+  "compress_ratios": [
+    128,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    0
+  ]
+}

config_org.json

@@ -0,0 +1,130 @@
+{
+  "architectures": [
+    "DeepseekV4ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "expert_dtype": "fp4",
+  "hc_eps": 1e-06,
+  "hc_mult": 4,
+  "hc_sinkhorn_iters": 20,
+  "head_dim": 512,
+  "hidden_act": "silu",
+  "hidden_size": 7168,
+  "index_head_dim": 128,
+  "index_n_heads": 64,
+  "index_topk": 1024,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 1048576,
+  "model_type": "deepseek_v4",
+  "moe_intermediate_size": 3072,
+  "n_routed_experts": 384,
+  "n_shared_experts": 1,
+  "norm_topk_prob": true,
+  "num_attention_heads": 128,
+  "num_experts_per_tok": 6,
+  "num_hidden_layers": 61,
+  "num_hash_layers": 3,
+  "num_key_value_heads": 1,
+  "num_nextn_predict_layers": 1,
+  "o_groups": 16,
+  "o_lora_rank": 1024,
+  "q_lora_rank": 1536,
+  "qk_rope_head_dim": 64,
+  "quantization_config": {
+    "activation_scheme": "dynamic",
+    "fmt": "e4m3",
+    "quant_method": "fp8",
+    "scale_fmt": "ue8m0",
+    "weight_block_size": [
+      128,
+      128
+    ]
+  },
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "beta_fast": 32,
+    "beta_slow": 1,
+    "factor": 16,
+    "original_max_position_embeddings": 65536,
+    "type": "yarn"
+  },
+  "rope_theta": 10000,
+  "routed_scaling_factor": 2.5,
+  "scoring_func": "sqrtsoftplus",
+  "sliding_window": 128,
+  "swiglu_limit": 10.0,
+  "tie_word_embeddings": false,
+  "topk_method": "noaux_tc",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "vocab_size": 129280,
+  "compress_rope_theta": 160000,
+  "compress_ratios": [
+    128,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    128,
+    4,
+    0
+  ]
+}

configuration.json

@@ -0,0 +1 @@
+{"framework":"Pytorch","task":"text-generation"}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "do_sample": true,
+  "temperature": 1.0,
+  "top_p": 1.0,
+  "transformers_version": "4.46.3"
+}

model-00001-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fa7abaaa9b64e833fe9a8e29b57d6dd93a6cd0f67533e8d18eafa9cb88182e9
+size 43997984936

model-00002-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec124afd4b6c05bf8f36caca50e9b2cfc10b82d63ef2a457fadab864b2de038a
+size 43964757308

model-00003-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fa16a7f6feb44dd4380659a1b5a8124183359a4dd91478b781827feb574d5f3
+size 43968700580

model-00004-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68f7a5e0eac30185f95bd1d75e5319cec4601c1ac0103bca3b8718753d1c5bb2
+size 43964757316

model-00005-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40571f0d227ba2d273418e0976e1393dec9f2dcdea35be5c891a16ecf8fe1378
+size 43962496660

model-00006-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d12fbe176bef41c6445b9412ef7a0b82c1322d9d84198f3542f2e1397b47a0a6
+size 43958553396

model-00007-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5dd84c05e13869cd0742413b67119bd78e8d5493ff12161c06f513b08fc1a6e
+size 43996265992

model-00008-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbf9831025ce8af1886c479086d78a90b4dc9b562b9055ac50723fbb2f133c4d
+size 43962496764

model-00009-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f9199d8cdae82514782ee94c95550a6a29e8a2f44ef9bb6839543bc5496d52a
+size 43958553396

model-00010-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c29613516d17237a070a2117c6b4ab4c8935b2aef949b69441f8d98be7568b0
+size 43962496660

model-00011-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f655ad9317fbac9d8aa33869fc10d295ba911a7e588e7b5e22995ada29edde39
+size 43958553396

model-00012-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5107be2abf242d1cefff653d91f9d862311243298f9d1a7d8a3a580c1c6bbb1
+size 43962496844

model-00013-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fe6704b9fb628b6f5274b72d2e8c5c89eb5dc5b5ca1c1b517ac939868abb14f
+size 43958554404

model-00014-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7d8170b79b357a221d858048fcbee71522f0df3b474dc149659e574a3746839
+size 43996267016

model-00015-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62d5a7461d90d24df20e45a50a7b54a6947a0ea2d0615145fb5e98e0e7d7397c
+size 43962497748

model-00016-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07d7d1d80ee8f7641f8acca160dae102fb2fc3e7f6894b4e2db1f138261cab9e
+size 43958554404

model-00017-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8db2dee87867c3d415ba319a1a829271505b2e410a414906a8f9856c4d0c1065
+size 43962497668

model-00018-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e5b089e5a3b6269ce0e2ca373b889e7e7809a64cae3c936ad600de64ba37f0c
+size 43958554404

model-00019-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caa27ba1eb8efe2b8c5526d7cf891cb099ab2aec3fda3e5c14c2fd05181b3e3a
+size 43962497668

model-00020-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acf3f0553f095b50e27d8ba3acb60d61a0b0fd6facf0b81bdab15e91ee6aabe3
+size 43958554404

model-00021-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81fc3a68b4c2492425dd464d0b1bc6f7be89287dca16a05163c1788bca67e770
+size 43996267048

model-00022-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf6b9b87e719d6553a753c390343307e69d8d8cd5d7d70cc36943ed38fb26270
+size 43962497716

model-00023-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c0dc2629ed6bbe28682d0331950c9e2eb79e67fe6fddd4fadcd31663aecc52c
+size 43958554404

model-00024-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db67b7bc2e753414b5f16dbce02407852c552b341564239b852599e660f4d84b
+size 43962497668

model-00025-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3d0199e2c8c2c11d27f6fff9079ea0bd07f93aea7eebd3e838f30ac14c376c5
+size 43958554404

model-00026-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ced4270fa0e2548e1f7537b38a3f81877de5583500d9fdd2f9b5cf24c7ba9b02
+size 43962497668

model-00027-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:963db657a385c845981cd85012f45fef44bf69562dbf01cc5c3657167626e901
+size 43933374508

model-00028-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b90f4ba53551a36d2ea0fae1450c158ba3abd1b99c1af6d73ce7ee3ab7c0d1d
+size 43977406672

model-00029-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22d8f499a35d3e5bcc89dce2946d97f52bdf8d2dd10797d1dbafcf567218a4da
+size 43962497684

model-00030-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53aa65024fb8c66c2f62b1b07f632f7a814280aacc83a7e7ce84bb37493b4bc7
+size 43958554412

model-00031-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa774e7876df661a898312fb1418d5a27289c142ae1557e0d223c7db61cf1851
+size 43962497660

model-00032-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:025ec48a59453c7ed4d0b1844ed94e449c6af64392519c8cca6915e38935beae
+size 43958554412

model-00033-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78ebaebb83e5504a029ef1eae2c482445547167a7f82c9b2794521f8b544b940
+size 43962497668

model-00034-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e08bf681ccfd8b94e2052cc87e29e41910a83182c7d37fb3349a2d6e5d60baa9
+size 43996266936

model-00035-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90e46ded70a0f6f55c12396b47e240f57b5b3c1ea7620513a3e4fbae99bccdb2
+size 43958554532

model-00036-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:312185f051863fee95666dc21a92ec378cf1563063a15b1dc7eedaa7d86d1417
+size 43962497660

model-00037-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ee3be8dc34c9e0c0544b2681acb2b8b653bc389e2a0ea98cfaa1a41b07c5ab2
+size 43958554412

model-00038-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7249b6c8ae840d593ca87f163d970051e563a63289b4ca00a2749fb5c15cb72e
+size 43962497660

model-00039-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2234e8924ba487f865f22de15688514ab4d558c4b2fbbd39294053cf73992d21
+size 43958554412

model-00040-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6903ef70daad7517c41857b39a2c3289f20b6b753fe7edafcba33b5faa24fe3
+size 43962497668

model-00041-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a2169bc3f2c7418cc3084cf16684ceac448683026f7ab09e9e888754f03aab7
+size 43996266968

model-00042-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a32de9df48fc020e3fcbabfb5a7c4caad0b976ead17c793872796e7189130f
+size 43958554548

model-00043-of-000073.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30d263d4b967b20f2ddc96bb3f9a88d56a6ef61be1a4d509cd30c5df4dd4d4d8
+size 43962497628