update

dnth · Oct 29, 2024 · 4f350fb · 4f350fb
1 parent fe9d43e
commit 4f350fb
Showing 1 changed file with 127 additions and 4 deletions.
diff --git a/nbs/phi35.ipynb b/nbs/phi35.ipynb
@@ -217,6 +217,77 @@
      },
      "metadata": {},
      "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "28fa86ac6e7843a4a8d0cd91795aafb3",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "model.safetensors.index.json:   0%|          | 0.00/68.9k [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "94c9459877284e38817ad2871b31ec37",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Loading safetensors checkpoint shards:   0% Completed | 0/2 [00:00<?, ?it/s]\n"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "INFO 10-29 23:54:28 model_runner.py:1067] Loading model weights took 7.9324 GB\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "8e15c49c8c2642af964bcf98fcd8bfa9",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "image_processing_phi3_v.py:   0%|          | 0.00/11.4k [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "A new version of the following files was downloaded from https://huggingface.co/microsoft/Phi-3-vision-128k-instruct:\n",
+      "- image_processing_phi3_v.py\n",
+      ". Make sure to double-check they do not contain any added malicious code. To avoid downloading new versions of the code file, you can pin a revision.\n",
+      "/home/dnth/mambaforge-pypy3/envs/xinfer/lib/python3.10/site-packages/transformers/models/auto/image_processing_auto.py:517: FutureWarning: The image_processor_class argument is deprecated and will be removed in v4.42. Please use `slow_image_processor_class`, or `fast_image_processor_class` instead\n",
+      "  warnings.warn(\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "INFO 10-29 23:54:30 gpu_executor.py:122] # GPU blocks: 2165, # CPU blocks: 682\n",
+      "INFO 10-29 23:54:30 gpu_executor.py:126] Maximum concurrency for 4096 tokens per request: 8.46x\n",
+      "INFO 10-29 23:54:32 model_runner.py:1395] Capturing the model for CUDA graphs. This may lead to unexpected consequences if the model is not static. To run the model in eager mode, set 'enforce_eager=True' or use '--enforce-eager' in the CLI.\n",
+      "INFO 10-29 23:54:32 model_runner.py:1399] CUDA graphs can take additional 1~3 GiB memory per GPU. If you are running out of memory, consider decreasing `gpu_memory_utilization` or enforcing eager mode. You can also reduce the `max_num_seqs` as needed to decrease memory usage.\n",
+      "INFO 10-29 23:54:33 model_runner.py:1523] Graph capturing finished in 1 secs.\n"
+     ]
     }
    ],
    "source": [
@@ -226,9 +297,34 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "prompt_token_ids (old) [1, 32010, 29871, 13, 29966, 29989, 3027, 29918, 29896, 29989, 29958, 13, 26270, 683, 445, 1967, 29889, 32007, 29871, 13, 32001]\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Processed prompts: 100%|██████████| 1/1 [00:00<00:00,  1.42it/s, est. speed input: 2742.73 toks/s, output: 22.68 toks/s]\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "'A woman with long hair and glasses is wearing a green blouse,'"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
    "source": [
     "image = \"../assets/demo/0a6ee446579d2885.jpg\"\n",
     "prompt = \"Caption this image.\"\n",
@@ -237,9 +333,36 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 4,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "prompt_token_ids (old) [1, 32010, 29871, 13, 29966, 29989, 3027, 29918, 29896, 29989, 29958, 13, 26270, 683, 445, 1967, 29889, 32007, 29871, 13, 32001]\n",
+      "prompt_token_ids (old) [1, 32010, 29871, 13, 29966, 29989, 3027, 29918, 29896, 29989, 29958, 13, 26270, 683, 445, 1967, 29889, 32007, 29871, 13, 32001]\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Processed prompts: 100%|██████████| 2/2 [00:01<00:00,  1.79it/s, est. speed input: 3466.46 toks/s, output: 28.66 toks/s]\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "['The image shows an individual with a blurred face, wearing a green',\n",
+       " 'A woman with light brown hair and glasses, wearing a green blouse']"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
    "source": [
     "model.infer_batch([image, image], [prompt, prompt])"
    ]