Update 2024-10-30 10:22:09

sgl-project · Oct 30, 2024 · a0e20e2 · a0e20e2
1 parent 64f1743
commit a0e20e2
Show file tree

Hide file tree

Showing 31 changed files with 1,280 additions and 799 deletions.
diff --git a/README.html b/README.html
@@ -174,6 +174,7 @@
 <li class="toctree-l1"><a class="reference internal" href="benchmark_and_profiling.html">Benchmark and Profiling</a></li>
 <li class="toctree-l1"><a class="reference internal" href="troubleshooting.html">Troubleshooting</a></li>
 <li class="toctree-l1"><a class="reference internal" href="embedding_model.html">Embedding Model</a></li>
+<li class="toctree-l1"><a class="reference internal" href="learn_more.html">Learn more</a></li>
 </ul>
 
     </div>

diff --git a/_sources/embedding_model.ipynb b/_sources/embedding_model.ipynb
@@ -32,52 +32,46 @@
    "execution_count": 1,
    "metadata": {
     "execution": {
-     "iopub.execute_input": "2024-10-30T09:50:44.977451Z",
-     "iopub.status.busy": "2024-10-30T09:50:44.977291Z",
-     "iopub.status.idle": "2024-10-30T09:51:32.305212Z",
-     "shell.execute_reply": "2024-10-30T09:51:32.304521Z"
+     "iopub.execute_input": "2024-10-30T10:18:30.036577Z",
+     "iopub.status.busy": "2024-10-30T10:18:30.036415Z",
+     "iopub.status.idle": "2024-10-30T10:19:14.359146Z",
+     "shell.execute_reply": "2024-10-30T10:19:14.358513Z"
     }
    },
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:50:54] server_args=ServerArgs(model_path='Alibaba-NLP/gte-Qwen2-7B-instruct', tokenizer_path='Alibaba-NLP/gte-Qwen2-7B-instruct', tokenizer_mode='auto', skip_tokenizer_init=False, load_format='auto', trust_remote_code=False, dtype='auto', kv_cache_dtype='auto', quantization=None, context_length=None, device='cuda', served_model_name='Alibaba-NLP/gte-Qwen2-7B-instruct', chat_template=None, is_embedding=True, host='0.0.0.0', port=30010, mem_fraction_static=0.88, max_running_requests=None, max_total_tokens=None, chunked_prefill_size=8192, max_prefill_tokens=16384, schedule_policy='lpm', schedule_conservativeness=1.0, tp_size=1, stream_interval=1, random_seed=448515216, constrained_json_whitespace_pattern=None, log_level='info', log_level_http=None, log_requests=False, show_time_cost=False, api_key=None, file_storage_pth='SGLang_storage', enable_cache_report=False, watchdog_timeout=600, dp_size=1, load_balance_method='round_robin', dist_init_addr=None, nnodes=1, node_rank=0, json_model_override_args='{}', enable_double_sparsity=False, ds_channel_config_path=None, ds_heavy_channel_num=32, ds_heavy_token_num=256, ds_heavy_channel_type='qk', ds_sparse_decode_threshold=4096, lora_paths=None, max_loras_per_batch=8, attention_backend='flashinfer', sampling_backend='flashinfer', grammar_backend='outlines', disable_flashinfer=False, disable_flashinfer_sampling=False, disable_radix_cache=False, disable_regex_jump_forward=False, disable_cuda_graph=False, disable_cuda_graph_padding=False, disable_disk_cache=False, disable_custom_all_reduce=False, disable_mla=False, disable_penalizer=False, disable_nan_detection=False, enable_overlap_schedule=False, enable_mixed_chunk=False, enable_torch_compile=False, torch_compile_max_bs=32, cuda_graph_max_bs=160, torchao_config='', enable_p2p_check=False, triton_attention_reduce_in_fp32=False, num_continuous_decode_steps=1)\n"
+      "[2024-10-30 10:18:40] server_args=ServerArgs(model_path='Alibaba-NLP/gte-Qwen2-7B-instruct', tokenizer_path='Alibaba-NLP/gte-Qwen2-7B-instruct', tokenizer_mode='auto', skip_tokenizer_init=False, load_format='auto', trust_remote_code=False, dtype='auto', kv_cache_dtype='auto', quantization=None, context_length=None, device='cuda', served_model_name='Alibaba-NLP/gte-Qwen2-7B-instruct', chat_template=None, is_embedding=True, host='0.0.0.0', port=30010, mem_fraction_static=0.88, max_running_requests=None, max_total_tokens=None, chunked_prefill_size=8192, max_prefill_tokens=16384, schedule_policy='lpm', schedule_conservativeness=1.0, tp_size=1, stream_interval=1, random_seed=922321184, constrained_json_whitespace_pattern=None, log_level='info', log_level_http=None, log_requests=False, show_time_cost=False, api_key=None, file_storage_pth='SGLang_storage', enable_cache_report=False, watchdog_timeout=600, dp_size=1, load_balance_method='round_robin', dist_init_addr=None, nnodes=1, node_rank=0, json_model_override_args='{}', enable_double_sparsity=False, ds_channel_config_path=None, ds_heavy_channel_num=32, ds_heavy_token_num=256, ds_heavy_channel_type='qk', ds_sparse_decode_threshold=4096, lora_paths=None, max_loras_per_batch=8, attention_backend='flashinfer', sampling_backend='flashinfer', grammar_backend='outlines', disable_flashinfer=False, disable_flashinfer_sampling=False, disable_radix_cache=False, disable_regex_jump_forward=False, disable_cuda_graph=False, disable_cuda_graph_padding=False, disable_disk_cache=False, disable_custom_all_reduce=False, disable_mla=False, disable_penalizer=False, disable_nan_detection=False, enable_overlap_schedule=False, enable_mixed_chunk=False, enable_torch_compile=False, torch_compile_max_bs=32, cuda_graph_max_bs=160, torchao_config='', enable_p2p_check=False, triton_attention_reduce_in_fp32=False, num_continuous_decode_steps=1)\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:10 TP0] Init torch distributed begin.\n"
+      "[2024-10-30 10:18:55 TP0] Init torch distributed begin.\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:11 TP0] Load weight begin. avail mem=78.59 GB\n"
+      "[2024-10-30 10:18:56 TP0] Load weight begin. avail mem=78.59 GB\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:12 TP0] lm_eval is not installed, GPTQ may not be usable\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "INFO 10-30 09:51:13 weight_utils.py:243] Using model weights format ['*.safetensors']\n"
+      "[2024-10-30 10:18:56 TP0] lm_eval is not installed, GPTQ may not be usable\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
+      "INFO 10-30 10:18:56 weight_utils.py:243] Using model weights format ['*.safetensors']\n",
       "\r",
       "Loading safetensors checkpoint shards:   0% Completed | 0/7 [00:00<?, ?it/s]\n"
      ]
@@ -87,106 +81,94 @@
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  14% Completed | 1/7 [00:01<00:10,  1.78s/it]\n"
+      "Loading safetensors checkpoint shards:  14% Completed | 1/7 [00:01<00:10,  1.68s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  29% Completed | 2/7 [00:03<00:09,  1.81s/it]\n"
+      "Loading safetensors checkpoint shards:  29% Completed | 2/7 [00:03<00:08,  1.76s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  43% Completed | 3/7 [00:05<00:07,  1.86s/it]\n"
+      "Loading safetensors checkpoint shards:  43% Completed | 3/7 [00:05<00:07,  1.82s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  57% Completed | 4/7 [00:07<00:05,  1.88s/it]\n"
+      "Loading safetensors checkpoint shards:  57% Completed | 4/7 [00:07<00:05,  1.85s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  71% Completed | 5/7 [00:09<00:03,  1.90s/it]\n"
+      "Loading safetensors checkpoint shards:  71% Completed | 5/7 [00:09<00:03,  1.83s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards:  86% Completed | 6/7 [00:10<00:01,  1.77s/it]\n"
+      "Loading safetensors checkpoint shards:  86% Completed | 6/7 [00:10<00:01,  1.67s/it]\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "\r",
-      "Loading safetensors checkpoint shards: 100% Completed | 7/7 [00:11<00:00,  1.47s/it]\n",
+      "Loading safetensors checkpoint shards: 100% Completed | 7/7 [00:11<00:00,  1.41s/it]\n",
       "\r",
-      "Loading safetensors checkpoint shards: 100% Completed | 7/7 [00:11<00:00,  1.68s/it]\n",
+      "Loading safetensors checkpoint shards: 100% Completed | 7/7 [00:11<00:00,  1.61s/it]\n",
       "\n",
-      "[2024-10-30 09:51:25 TP0] Load weight end. type=Qwen2ForCausalLM, dtype=torch.float16, avail mem=64.18 GB\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "[2024-10-30 09:51:25 TP0] Memory pool end. avail mem=7.43 GB\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "[2024-10-30 09:51:26 TP0] max_total_num_tokens=1025173, max_prefill_tokens=16384, max_running_requests=4005, context_len=131072\n"
+      "[2024-10-30 10:19:08 TP0] Load weight end. type=Qwen2ForCausalLM, dtype=torch.float16, avail mem=64.18 GB\n",
+      "[2024-10-30 10:19:08 TP0] Memory pool end. avail mem=7.43 GB\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:26] INFO:     Started server process [1244882]\n",
-      "[2024-10-30 09:51:26] INFO:     Waiting for application startup.\n",
-      "[2024-10-30 09:51:26] INFO:     Application startup complete.\n",
-      "[2024-10-30 09:51:26] INFO:     Uvicorn running on http://0.0.0.0:30010 (Press CTRL+C to quit)\n"
+      "[2024-10-30 10:19:08 TP0] max_total_num_tokens=1025173, max_prefill_tokens=16384, max_running_requests=4005, context_len=131072\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:27] INFO:     127.0.0.1:43056 - \"GET /v1/models HTTP/1.1\" 200 OK\n"
+      "[2024-10-30 10:19:09] INFO:     Started server process [2218395]\n",
+      "[2024-10-30 10:19:09] INFO:     Waiting for application startup.\n",
+      "[2024-10-30 10:19:09] INFO:     Application startup complete.\n",
+      "[2024-10-30 10:19:09] INFO:     Uvicorn running on http://0.0.0.0:30010 (Press CTRL+C to quit)\n",
+      "[2024-10-30 10:19:09] INFO:     127.0.0.1:49928 - \"GET /v1/models HTTP/1.1\" 200 OK\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:27] INFO:     127.0.0.1:43062 - \"GET /get_model_info HTTP/1.1\" 200 OK\n",
-      "[2024-10-30 09:51:27 TP0] Prefill batch. #new-seq: 1, #new-token: 6, #cached-token: 0, cache hit rate: 0.00%, token usage: 0.00, #running-req: 0, #queue-req: 0\n"
+      "[2024-10-30 10:19:10] INFO:     127.0.0.1:49940 - \"GET /get_model_info HTTP/1.1\" 200 OK\n",
+      "[2024-10-30 10:19:10 TP0] Prefill batch. #new-seq: 1, #new-token: 6, #cached-token: 0, cache hit rate: 0.00%, token usage: 0.00, #running-req: 0, #queue-req: 0\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:28] INFO:     127.0.0.1:43078 - \"POST /encode HTTP/1.1\" 200 OK\n",
-      "[2024-10-30 09:51:28] The server is fired up and ready to roll!\n"
+      "[2024-10-30 10:19:11] INFO:     127.0.0.1:49946 - \"POST /encode HTTP/1.1\" 200 OK\n",
+      "[2024-10-30 10:19:11] The server is fired up and ready to roll!\n"
      ]
     },
     {
@@ -232,19 +214,19 @@
    "execution_count": 2,
    "metadata": {
     "execution": {
-     "iopub.execute_input": "2024-10-30T09:51:32.307542Z",
-     "iopub.status.busy": "2024-10-30T09:51:32.307258Z",
-     "iopub.status.idle": "2024-10-30T09:51:32.349027Z",
-     "shell.execute_reply": "2024-10-30T09:51:32.348350Z"
+     "iopub.execute_input": "2024-10-30T10:19:14.361455Z",
+     "iopub.status.busy": "2024-10-30T10:19:14.361206Z",
+     "iopub.status.idle": "2024-10-30T10:19:14.399746Z",
+     "shell.execute_reply": "2024-10-30T10:19:14.399141Z"
     }
    },
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:32 TP0] Prefill batch. #new-seq: 1, #new-token: 4, #cached-token: 0, cache hit rate: 0.00%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
-      "[2024-10-30 09:51:32] INFO:     127.0.0.1:34504 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
+      "[2024-10-30 10:19:14 TP0] Prefill batch. #new-seq: 1, #new-token: 4, #cached-token: 0, cache hit rate: 0.00%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
+      "[2024-10-30 10:19:14] INFO:     127.0.0.1:49980 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
      ]
     },
     {
@@ -289,19 +271,19 @@
    "execution_count": 3,
    "metadata": {
     "execution": {
-     "iopub.execute_input": "2024-10-30T09:51:32.350944Z",
-     "iopub.status.busy": "2024-10-30T09:51:32.350735Z",
-     "iopub.status.idle": "2024-10-30T09:51:32.816553Z",
-     "shell.execute_reply": "2024-10-30T09:51:32.815847Z"
+     "iopub.execute_input": "2024-10-30T10:19:14.401590Z",
+     "iopub.status.busy": "2024-10-30T10:19:14.401411Z",
+     "iopub.status.idle": "2024-10-30T10:19:14.850015Z",
+     "shell.execute_reply": "2024-10-30T10:19:14.849423Z"
     }
    },
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:32 TP0] Prefill batch. #new-seq: 1, #new-token: 1, #cached-token: 3, cache hit rate: 21.43%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
-      "[2024-10-30 09:51:32] INFO:     127.0.0.1:34520 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
+      "[2024-10-30 10:19:14 TP0] Prefill batch. #new-seq: 1, #new-token: 1, #cached-token: 3, cache hit rate: 21.43%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
+      "[2024-10-30 10:19:14] INFO:     127.0.0.1:49986 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
      ]
     },
     {
@@ -346,19 +328,19 @@
    "execution_count": 4,
    "metadata": {
     "execution": {
-     "iopub.execute_input": "2024-10-30T09:51:32.818545Z",
-     "iopub.status.busy": "2024-10-30T09:51:32.818328Z",
-     "iopub.status.idle": "2024-10-30T09:51:39.897939Z",
-     "shell.execute_reply": "2024-10-30T09:51:39.897207Z"
+     "iopub.execute_input": "2024-10-30T10:19:14.851882Z",
+     "iopub.status.busy": "2024-10-30T10:19:14.851691Z",
+     "iopub.status.idle": "2024-10-30T10:19:21.910313Z",
+     "shell.execute_reply": "2024-10-30T10:19:21.909605Z"
     }
    },
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[2024-10-30 09:51:39 TP0] Prefill batch. #new-seq: 1, #new-token: 1, #cached-token: 3, cache hit rate: 33.33%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
-      "[2024-10-30 09:51:39] INFO:     127.0.0.1:42628 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
+      "[2024-10-30 10:19:21 TP0] Prefill batch. #new-seq: 1, #new-token: 1, #cached-token: 3, cache hit rate: 33.33%, token usage: 0.00, #running-req: 0, #queue-req: 0\n",
+      "[2024-10-30 10:19:21] INFO:     127.0.0.1:43896 - \"POST /v1/embeddings HTTP/1.1\" 200 OK\n"
      ]
     },
     {
@@ -401,10 +383,10 @@
    "execution_count": 5,
    "metadata": {
     "execution": {
-     "iopub.execute_input": "2024-10-30T09:51:39.900119Z",
-     "iopub.status.busy": "2024-10-30T09:51:39.899764Z",
-     "iopub.status.idle": "2024-10-30T09:51:40.354387Z",
-     "shell.execute_reply": "2024-10-30T09:51:40.353699Z"
+     "iopub.execute_input": "2024-10-30T10:19:21.912691Z",
+     "iopub.status.busy": "2024-10-30T10:19:21.912221Z",
+     "iopub.status.idle": "2024-10-30T10:19:22.397961Z",
+     "shell.execute_reply": "2024-10-30T10:19:22.397219Z"
     }
    },
    "outputs": [],

diff --git a/_sources/index.rst b/_sources/index.rst
@@ -46,3 +46,4 @@ The core features include:
    benchmark_and_profiling.md
    troubleshooting.md
    embedding_model.ipynb
+   learn_more.md
diff --git a/_sources/learn_more.md b/_sources/learn_more.md
@@ -0,0 +1,3 @@
+# Learn more
+
+You can find more blogs, slides, and videos about SGLang at [https://github.com/sgl-project/sgl-learning-materials](https://github.com/sgl-project/sgl-learning-materials).