Devops-hestabit
/

mixtral-instruct-trt

Model card Files Files and versions Community

Devops-hestabit commited on Jul 1, 2024

Commit

ec365ce

verified ·

1 Parent(s): 4f10fb0

Upload folder using huggingface_hub

Browse files

Files changed (8) hide show

postprocessing/1/__pycache__/model.cpython-310.pyc +0 -0
preprocessing/1/__pycache__/model.cpython-310.pyc +0 -0
tensorrt_llm/1/config.json +5 -5
tensorrt_llm/1/rank0.engine +2 -2
tensorrt_llm/1/rank1.engine +2 -2
tensorrt_llm_bls/1/__pycache__/model.cpython-310.pyc +0 -0
tensorrt_llm_bls/1/lib/__pycache__/decode.cpython-310.pyc +0 -0
tensorrt_llm_bls/1/lib/__pycache__/triton_decoder.cpython-310.pyc +0 -0

postprocessing/1/__pycache__/model.cpython-310.pyc CHANGED Viewed

Binary files a/postprocessing/1/__pycache__/model.cpython-310.pyc and b/postprocessing/1/__pycache__/model.cpython-310.pyc differ

preprocessing/1/__pycache__/model.cpython-310.pyc CHANGED Viewed

Binary files a/preprocessing/1/__pycache__/model.cpython-310.pyc and b/preprocessing/1/__pycache__/model.cpython-310.pyc differ

tensorrt_llm/1/config.json CHANGED Viewed

@@ -28,9 +28,9 @@
         "mapping": {
             "world_size": 2,
             "gpus_per_node": 8,
-            "tp_size": 1,
-            "pp_size": 2,
-            "moe_tp_size": 1,
             "moe_ep_size": 1
         },
         "quantization": {
@@ -50,11 +50,11 @@
     },
     "build_config": {
         "max_input_len": 28000,
-        "max_seq_len": 32500,
         "opt_batch_size": null,
         "max_batch_size": 16,
         "max_beam_width": 1,
-        "max_num_tokens": 8192,
         "opt_num_tokens": 16,
         "max_prompt_embedding_table_size": 0,
         "gather_context_logits": false,

         "mapping": {
             "world_size": 2,
             "gpus_per_node": 8,
+            "tp_size": 2,
+            "pp_size": 1,
+            "moe_tp_size": 2,
             "moe_ep_size": 1
         },
         "quantization": {
     },
     "build_config": {
         "max_input_len": 28000,
+        "max_seq_len": 32000,
         "opt_batch_size": null,
         "max_batch_size": 16,
         "max_beam_width": 1,
+        "max_num_tokens": 32000,
         "opt_num_tokens": 16,
         "max_prompt_embedding_table_size": 0,
         "gather_context_logits": false,

tensorrt_llm/1/rank0.engine CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7acd8fbed6cb1b6373e7e00f78cb5910ce6deeb7f5f606ad6cb2123540049bf9
-size 46722172788

 version https://git-lfs.github.com/spec/v1
+oid sha256:b03d07bc220139d071a7d50dd42a7c67235601b1d28a2aea79db0f65ea8150b4
+size 46857284148

tensorrt_llm/1/rank1.engine CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1e59d5a7bb7d8d7c87dcc81f50518d6cb7d6a0e4fde1f60ac194e02fb4ac86e
-size 46722181404

 version https://git-lfs.github.com/spec/v1
+oid sha256:1397dbb8fc710907448a5c58da6c17004b8b8fa398780b1e593b383e6b150b6f
+size 46857271996

tensorrt_llm_bls/1/__pycache__/model.cpython-310.pyc CHANGED Viewed

Binary files a/tensorrt_llm_bls/1/__pycache__/model.cpython-310.pyc and b/tensorrt_llm_bls/1/__pycache__/model.cpython-310.pyc differ

tensorrt_llm_bls/1/lib/__pycache__/decode.cpython-310.pyc CHANGED Viewed

Binary files a/tensorrt_llm_bls/1/lib/__pycache__/decode.cpython-310.pyc and b/tensorrt_llm_bls/1/lib/__pycache__/decode.cpython-310.pyc differ

tensorrt_llm_bls/1/lib/__pycache__/triton_decoder.cpython-310.pyc CHANGED Viewed

Binary files a/tensorrt_llm_bls/1/lib/__pycache__/triton_decoder.cpython-310.pyc and b/tensorrt_llm_bls/1/lib/__pycache__/triton_decoder.cpython-310.pyc differ