Add support for GPTQ Marlin

Dockerfile

@@ -140,9 +140,9 @@ RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-eetq
 # Build marlin kernels
 FROM kernel-builder as marlin-kernels-builder
 WORKDIR /usr/src
-COPY server/Makefile-marlin Makefile
+COPY server/marlin/ .
 # Build specific version of transformers
-RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-marlin
+RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" python setup.py build
 
 # Build Transformers CUDA kernels
 FROM kernel-builder as custom-kernels-builder
@@ -213,7 +213,7 @@ COPY --from=awq-kernels-builder /usr/src/llm-awq/awq/kernels/build/lib.linux-x86
 # Copy build artifacts from eetq kernels builder
 COPY --from=eetq-kernels-builder /usr/src/eetq/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
 # Copy build artifacts from marlin kernels builder
-COPY --from=marlin-kernels-builder /usr/src/marlin/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
+COPY --from=marlin-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
 
 # Copy builds artifacts from vllm builder
 COPY --from=vllm-builder /usr/src/vllm/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages

integration-tests/models/__snapshots__/test_flash_llama_gptq_marlin/test_flash_llama_gptq_marlin.json

@@ -0,0 +1,84 @@
+{
+  "details": {
+    "best_of_sequences": null,
+    "finish_reason": "length",
+    "generated_tokens": 10,
+    "prefill": [
+      {
+        "id": 2323,
+        "logprob": null,
+        "text": "Test"
+      },
+      {
+        "id": 1715,
+        "logprob": -11.75,
+        "text": " request"
+      }
+    ],
+    "seed": null,
+    "tokens": [
+      {
+        "id": 311,
+        "logprob": -2.8261719,
+        "special": false,
+        "text": " to"
+      },
+      {
+        "id": 279,
+        "logprob": -2.0507812,
+        "special": false,
+        "text": " the"
+      },
+      {
+        "id": 3622,
+        "logprob": -2.2832031,
+        "special": false,
+        "text": " server"
+      },
+      {
+        "id": 13,
+        "logprob": -1.7314453,
+        "special": false,
+        "text": "."
+      },
+      {
+        "id": 578,
+        "logprob": -1.5800781,
+        "special": false,
+        "text": " The"
+      },
+      {
+        "id": 1715,
+        "logprob": -1.3330078,
+        "special": false,
+        "text": " request"
+      },
+      {
+        "id": 374,
+        "logprob": -1.4277344,
+        "special": false,
+        "text": " is"
+      },
+      {
+        "id": 3288,
+        "logprob": -1.8115234,
+        "special": false,
+        "text": " sent"
+      },
+      {
+        "id": 311,
+        "logprob": -1.1923828,
+        "special": false,
+        "text": " to"
+      },
+      {
+        "id": 279,
+        "logprob": -0.06713867,
+        "special": false,
+        "text": " the"
+      }
+    ],
+    "top_tokens": null
+  },
+  "generated_text": " to the server. The request is sent to the"
+}

integration-tests/models/__snapshots__/test_flash_llama_gptq_marlin/test_flash_llama_gptq_marlin_all_params.json

@@ -0,0 +1,84 @@
+{
+  "details": {
+    "best_of_sequences": null,
+    "finish_reason": "length",
+    "generated_tokens": 10,
+    "prefill": [
+      {
+        "id": 2323,
+        "logprob": null,
+        "text": "Test"
+      },
+      {
+        "id": 1715,
+        "logprob": -11.75,
+        "text": " request"
+      }
+    ],
+    "seed": 0,
+    "tokens": [
+      {
+        "id": 13,
+        "logprob": -2.2363281,
+        "special": false,
+        "text": "."
+      },
+      {
+        "id": 578,
+        "logprob": -0.10021973,
+        "special": false,
+        "text": " The"
+      },
+      {
+        "id": 3016,
+        "logprob": -1.6083984,
+        "special": false,
+        "text": " client"
+      },
+      {
+        "id": 649,
+        "logprob": -0.95751953,
+        "special": false,
+        "text": " can"
+      },
+      {
+        "id": 1005,
+        "logprob": -0.6298828,
+        "special": false,
+        "text": " use"
+      },
+      {
+        "id": 279,
+        "logprob": 0.0,
+        "special": false,
+        "text": " the"
+      },
+      {
+        "id": 1595,
+        "logprob": -0.17492676,
+        "special": false,
+        "text": " `"
+      },
+      {
+        "id": 2079,
+        "logprob": -0.27368164,
+        "special": false,
+        "text": "request"
+      },
+      {
+        "id": 63,
+        "logprob": 0.0,
+        "special": false,
+        "text": "`"
+      },
+      {
+        "id": 1665,
+        "logprob": 0.0,
+        "special": false,
+        "text": " object"
+      }
+    ],
+    "top_tokens": null
+  },
+  "generated_text": "Test request. The client can use the `request` object"
+}