Spaces:
Running
Running
fix: typo
Browse files
dev/inference/wandb-backend.ipynb
CHANGED
@@ -2,7 +2,7 @@
|
|
2 |
"cells": [
|
3 |
{
|
4 |
"cell_type": "code",
|
5 |
-
"execution_count":
|
6 |
"id": "4ff2a984-b8b2-4a69-89cf-0d16da2393c8",
|
7 |
"metadata": {},
|
8 |
"outputs": [],
|
@@ -12,7 +12,7 @@
|
|
12 |
"import random\n",
|
13 |
"import numpy as np\n",
|
14 |
"from PIL import Image\n",
|
15 |
-
"from tqdm import tqdm\n",
|
16 |
"import jax\n",
|
17 |
"import jax.numpy as jnp\n",
|
18 |
"from flax.training.common_utils import shard, shard_prng_key\n",
|
@@ -26,7 +26,7 @@
|
|
26 |
},
|
27 |
{
|
28 |
"cell_type": "code",
|
29 |
-
"execution_count":
|
30 |
"id": "92f4557c-fd7f-4edc-81c2-de0b0a10c270",
|
31 |
"metadata": {},
|
32 |
"outputs": [],
|
@@ -36,13 +36,13 @@
|
|
36 |
"VQGAN_REPO, VQGAN_COMMIT_ID = 'dalle-mini/vqgan_imagenet_f16_16384', None\n",
|
37 |
"normalize_text = True\n",
|
38 |
"latest_only = False # log only latest or all versions\n",
|
39 |
-
"suffix = '
|
40 |
"add_clip_32 = False"
|
41 |
]
|
42 |
},
|
43 |
{
|
44 |
"cell_type": "code",
|
45 |
-
"execution_count":
|
46 |
"id": "23e00271-941c-4e1b-b6a9-107a1b77324d",
|
47 |
"metadata": {},
|
48 |
"outputs": [],
|
@@ -52,16 +52,25 @@
|
|
52 |
"VQGAN_REPO, VQGAN_COMMIT_ID = 'dalle-mini/vqgan_imagenet_f16_16384', None\n",
|
53 |
"normalize_text = False\n",
|
54 |
"latest_only = True # log only latest or all versions\n",
|
55 |
-
"suffix = '
|
56 |
"add_clip_32 = True"
|
57 |
]
|
58 |
},
|
59 |
{
|
60 |
"cell_type": "code",
|
61 |
-
"execution_count":
|
62 |
"id": "93b2e24b-f0e5-4abe-a3ec-0aa834cc3bf3",
|
63 |
"metadata": {},
|
64 |
-
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
"source": [
|
66 |
"batch_size = 8\n",
|
67 |
"num_images = 128\n",
|
@@ -75,10 +84,18 @@
|
|
75 |
},
|
76 |
{
|
77 |
"cell_type": "code",
|
78 |
-
"execution_count":
|
79 |
"id": "c6a878fa-4bf5-4978-abb5-e235841d765b",
|
80 |
"metadata": {},
|
81 |
-
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
"source": [
|
83 |
"vqgan = VQModel.from_pretrained(VQGAN_REPO, revision=VQGAN_COMMIT_ID)\n",
|
84 |
"clip = FlaxCLIPModel.from_pretrained(\"openai/clip-vit-base-patch16\")\n",
|
@@ -94,7 +111,7 @@
|
|
94 |
},
|
95 |
{
|
96 |
"cell_type": "code",
|
97 |
-
"execution_count":
|
98 |
"id": "a500dd07-dbc3-477d-80d4-2b73a3b83ef3",
|
99 |
"metadata": {},
|
100 |
"outputs": [],
|
@@ -104,20 +121,42 @@
|
|
104 |
" return vqgan.decode_code(indices, params=params)\n",
|
105 |
"\n",
|
106 |
"@partial(jax.pmap, axis_name=\"batch\")\n",
|
107 |
-
"def p_clip(inputs):\n",
|
108 |
-
" logits = clip(params=
|
109 |
" return logits\n",
|
110 |
"\n",
|
111 |
"if add_clip_32:\n",
|
112 |
" @partial(jax.pmap, axis_name=\"batch\")\n",
|
113 |
-
" def p_clip32(inputs):\n",
|
114 |
-
" logits = clip32(params=
|
115 |
" return logits"
|
116 |
]
|
117 |
},
|
118 |
{
|
119 |
"cell_type": "code",
|
120 |
-
"execution_count":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
121 |
"id": "e57797ab-0b3a-4490-be58-03d8d1c23fe9",
|
122 |
"metadata": {},
|
123 |
"outputs": [],
|
@@ -133,7 +172,7 @@
|
|
133 |
},
|
134 |
{
|
135 |
"cell_type": "code",
|
136 |
-
"execution_count":
|
137 |
"id": "f3e02d9d-4ee1-49e7-a7bc-4d8b139e9614",
|
138 |
"metadata": {},
|
139 |
"outputs": [],
|
@@ -150,7 +189,7 @@
|
|
150 |
},
|
151 |
{
|
152 |
"cell_type": "code",
|
153 |
-
"execution_count":
|
154 |
"id": "f0d7ed17-7abb-4a31-ab3c-a12b9039a570",
|
155 |
"metadata": {},
|
156 |
"outputs": [],
|
@@ -163,7 +202,7 @@
|
|
163 |
},
|
164 |
{
|
165 |
"cell_type": "code",
|
166 |
-
"execution_count":
|
167 |
"id": "7e784a43-626d-4e8d-9e47-a23775b2f35f",
|
168 |
"metadata": {},
|
169 |
"outputs": [],
|
@@ -179,7 +218,7 @@
|
|
179 |
},
|
180 |
{
|
181 |
"cell_type": "code",
|
182 |
-
"execution_count":
|
183 |
"id": "d1cc9993-1bfc-4ec6-a004-c056189c42ac",
|
184 |
"metadata": {},
|
185 |
"outputs": [],
|
@@ -202,7 +241,7 @@
|
|
202 |
},
|
203 |
{
|
204 |
"cell_type": "code",
|
205 |
-
"execution_count":
|
206 |
"id": "23b2444c-67a9-44d7-abd1-187ed83a9431",
|
207 |
"metadata": {},
|
208 |
"outputs": [],
|
@@ -213,10 +252,19 @@
|
|
213 |
},
|
214 |
{
|
215 |
"cell_type": "code",
|
216 |
-
"execution_count":
|
217 |
"id": "bba70f33-af8b-4eb3-9973-7be672301a0b",
|
218 |
"metadata": {},
|
219 |
-
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
220 |
"source": [
|
221 |
"artifact_versions = get_artifact_versions(run_id, latest_only)\n",
|
222 |
"last_inference_version = get_last_inference_version(run_id)\n",
|
@@ -276,9 +324,8 @@
|
|
276 |
" tokenized_prompt = shard(tokenized_prompt)\n",
|
277 |
"\n",
|
278 |
" # generate images\n",
|
279 |
-
" print('Generating images')\n",
|
280 |
" images = []\n",
|
281 |
-
" for i in tqdm(range(num_images // jax.device_count())):\n",
|
282 |
" key, subkey = jax.random.split(key)\n",
|
283 |
" encoded_images = p_generate(tokenized_prompt, shard_prng_key(subkey), model_params)\n",
|
284 |
" encoded_images = encoded_images.sequences[..., 1:]\n",
|
@@ -294,7 +341,7 @@
|
|
294 |
" images_per_prompt_indices = np.asarray(range(0, len(images), batch_size))\n",
|
295 |
" clip_inputs['pixel_values'] = jnp.concatenate(list(clip_inputs['pixel_values'][images_per_prompt_indices + i] for i in range(batch_size)))\n",
|
296 |
" clip_inputs = shard(clip_inputs)\n",
|
297 |
-
" logits = p_clip(clip_inputs)\n",
|
298 |
" logits = logits.reshape(-1, num_images)\n",
|
299 |
" top_scores = logits.argsort()[:, -top_k:][..., ::-1]\n",
|
300 |
" logits = jax.device_get(logits)\n",
|
@@ -314,7 +361,7 @@
|
|
314 |
" images_per_prompt_indices = np.asarray(range(0, len(images), batch_size))\n",
|
315 |
" clip_inputs['pixel_values'] = jnp.concatenate(list(clip_inputs['pixel_values'][images_per_prompt_indices + i] for i in range(batch_size)))\n",
|
316 |
" clip_inputs = shard(clip_inputs)\n",
|
317 |
-
" logits = p_clip32(clip_inputs)\n",
|
318 |
" logits = logits.reshape(-1, num_images)\n",
|
319 |
" top_scores = logits.argsort()[:, -top_k:][..., ::-1]\n",
|
320 |
" logits = jax.device_get(logits)\n",
|
|
|
2 |
"cells": [
|
3 |
{
|
4 |
"cell_type": "code",
|
5 |
+
"execution_count": 1,
|
6 |
"id": "4ff2a984-b8b2-4a69-89cf-0d16da2393c8",
|
7 |
"metadata": {},
|
8 |
"outputs": [],
|
|
|
12 |
"import random\n",
|
13 |
"import numpy as np\n",
|
14 |
"from PIL import Image\n",
|
15 |
+
"from tqdm.notebook import tqdm\n",
|
16 |
"import jax\n",
|
17 |
"import jax.numpy as jnp\n",
|
18 |
"from flax.training.common_utils import shard, shard_prng_key\n",
|
|
|
26 |
},
|
27 |
{
|
28 |
"cell_type": "code",
|
29 |
+
"execution_count": 2,
|
30 |
"id": "92f4557c-fd7f-4edc-81c2-de0b0a10c270",
|
31 |
"metadata": {},
|
32 |
"outputs": [],
|
|
|
36 |
"VQGAN_REPO, VQGAN_COMMIT_ID = 'dalle-mini/vqgan_imagenet_f16_16384', None\n",
|
37 |
"normalize_text = True\n",
|
38 |
"latest_only = False # log only latest or all versions\n",
|
39 |
+
"suffix = '' # mainly for duplicate inference runs with a deleted version\n",
|
40 |
"add_clip_32 = False"
|
41 |
]
|
42 |
},
|
43 |
{
|
44 |
"cell_type": "code",
|
45 |
+
"execution_count": 3,
|
46 |
"id": "23e00271-941c-4e1b-b6a9-107a1b77324d",
|
47 |
"metadata": {},
|
48 |
"outputs": [],
|
|
|
52 |
"VQGAN_REPO, VQGAN_COMMIT_ID = 'dalle-mini/vqgan_imagenet_f16_16384', None\n",
|
53 |
"normalize_text = False\n",
|
54 |
"latest_only = True # log only latest or all versions\n",
|
55 |
+
"suffix = '' # mainly for duplicate inference runs with a deleted version\n",
|
56 |
"add_clip_32 = True"
|
57 |
]
|
58 |
},
|
59 |
{
|
60 |
"cell_type": "code",
|
61 |
+
"execution_count": 4,
|
62 |
"id": "93b2e24b-f0e5-4abe-a3ec-0aa834cc3bf3",
|
63 |
"metadata": {},
|
64 |
+
"outputs": [
|
65 |
+
{
|
66 |
+
"name": "stderr",
|
67 |
+
"output_type": "stream",
|
68 |
+
"text": [
|
69 |
+
"INFO:absl:Unable to initialize backend 'tpu_driver': NOT_FOUND: Unable to find driver in registry given worker: \n",
|
70 |
+
"INFO:absl:Unable to initialize backend 'gpu': NOT_FOUND: Could not find registered platform with name: \"cuda\". Available platform names are: TPU Interpreter Host\n"
|
71 |
+
]
|
72 |
+
}
|
73 |
+
],
|
74 |
"source": [
|
75 |
"batch_size = 8\n",
|
76 |
"num_images = 128\n",
|
|
|
84 |
},
|
85 |
{
|
86 |
"cell_type": "code",
|
87 |
+
"execution_count": 5,
|
88 |
"id": "c6a878fa-4bf5-4978-abb5-e235841d765b",
|
89 |
"metadata": {},
|
90 |
+
"outputs": [
|
91 |
+
{
|
92 |
+
"name": "stdout",
|
93 |
+
"output_type": "stream",
|
94 |
+
"text": [
|
95 |
+
"Working with z of shape (1, 256, 16, 16) = 65536 dimensions.\n"
|
96 |
+
]
|
97 |
+
}
|
98 |
+
],
|
99 |
"source": [
|
100 |
"vqgan = VQModel.from_pretrained(VQGAN_REPO, revision=VQGAN_COMMIT_ID)\n",
|
101 |
"clip = FlaxCLIPModel.from_pretrained(\"openai/clip-vit-base-patch16\")\n",
|
|
|
111 |
},
|
112 |
{
|
113 |
"cell_type": "code",
|
114 |
+
"execution_count": 6,
|
115 |
"id": "a500dd07-dbc3-477d-80d4-2b73a3b83ef3",
|
116 |
"metadata": {},
|
117 |
"outputs": [],
|
|
|
121 |
" return vqgan.decode_code(indices, params=params)\n",
|
122 |
"\n",
|
123 |
"@partial(jax.pmap, axis_name=\"batch\")\n",
|
124 |
+
"def p_clip(inputs, params):\n",
|
125 |
+
" logits = clip(params=params, **inputs).logits_per_image\n",
|
126 |
" return logits\n",
|
127 |
"\n",
|
128 |
"if add_clip_32:\n",
|
129 |
" @partial(jax.pmap, axis_name=\"batch\")\n",
|
130 |
+
" def p_clip32(inputs, params):\n",
|
131 |
+
" logits = clip32(params=params, **inputs).logits_per_image\n",
|
132 |
" return logits"
|
133 |
]
|
134 |
},
|
135 |
{
|
136 |
"cell_type": "code",
|
137 |
+
"execution_count": 7,
|
138 |
+
"id": "ebf4f7bf-2efa-46cc-b3f4-2d7a54f7b2cb",
|
139 |
+
"metadata": {},
|
140 |
+
"outputs": [
|
141 |
+
{
|
142 |
+
"data": {
|
143 |
+
"text/plain": [
|
144 |
+
"ShardedDeviceArray([4.6051702, 4.6051702, 4.6051702, 4.6051702, 4.6051702,\n",
|
145 |
+
" 4.6051702, 4.6051702, 4.6051702], dtype=float32)"
|
146 |
+
]
|
147 |
+
},
|
148 |
+
"execution_count": 7,
|
149 |
+
"metadata": {},
|
150 |
+
"output_type": "execute_result"
|
151 |
+
}
|
152 |
+
],
|
153 |
+
"source": [
|
154 |
+
"clip_params['logit_scale']"
|
155 |
+
]
|
156 |
+
},
|
157 |
+
{
|
158 |
+
"cell_type": "code",
|
159 |
+
"execution_count": 8,
|
160 |
"id": "e57797ab-0b3a-4490-be58-03d8d1c23fe9",
|
161 |
"metadata": {},
|
162 |
"outputs": [],
|
|
|
172 |
},
|
173 |
{
|
174 |
"cell_type": "code",
|
175 |
+
"execution_count": 9,
|
176 |
"id": "f3e02d9d-4ee1-49e7-a7bc-4d8b139e9614",
|
177 |
"metadata": {},
|
178 |
"outputs": [],
|
|
|
189 |
},
|
190 |
{
|
191 |
"cell_type": "code",
|
192 |
+
"execution_count": 10,
|
193 |
"id": "f0d7ed17-7abb-4a31-ab3c-a12b9039a570",
|
194 |
"metadata": {},
|
195 |
"outputs": [],
|
|
|
202 |
},
|
203 |
{
|
204 |
"cell_type": "code",
|
205 |
+
"execution_count": 11,
|
206 |
"id": "7e784a43-626d-4e8d-9e47-a23775b2f35f",
|
207 |
"metadata": {},
|
208 |
"outputs": [],
|
|
|
218 |
},
|
219 |
{
|
220 |
"cell_type": "code",
|
221 |
+
"execution_count": 12,
|
222 |
"id": "d1cc9993-1bfc-4ec6-a004-c056189c42ac",
|
223 |
"metadata": {},
|
224 |
"outputs": [],
|
|
|
241 |
},
|
242 |
{
|
243 |
"cell_type": "code",
|
244 |
+
"execution_count": 13,
|
245 |
"id": "23b2444c-67a9-44d7-abd1-187ed83a9431",
|
246 |
"metadata": {},
|
247 |
"outputs": [],
|
|
|
252 |
},
|
253 |
{
|
254 |
"cell_type": "code",
|
255 |
+
"execution_count": 14,
|
256 |
"id": "bba70f33-af8b-4eb3-9973-7be672301a0b",
|
257 |
"metadata": {},
|
258 |
+
"outputs": [
|
259 |
+
{
|
260 |
+
"ename": "SyntaxError",
|
261 |
+
"evalue": "EOL while scanning string literal (1745443972.py, line 60)",
|
262 |
+
"output_type": "error",
|
263 |
+
"traceback": [
|
264 |
+
"\u001b[0;36m File \u001b[0;32m\"/tmp/ipykernel_402605/1745443972.py\"\u001b[0;36m, line \u001b[0;32m60\u001b[0m\n\u001b[0;31m for i in tqdm(range(num_images // jax.device_count()), desc='Generating Images):\u001b[0m\n\u001b[0m ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m EOL while scanning string literal\n"
|
265 |
+
]
|
266 |
+
}
|
267 |
+
],
|
268 |
"source": [
|
269 |
"artifact_versions = get_artifact_versions(run_id, latest_only)\n",
|
270 |
"last_inference_version = get_last_inference_version(run_id)\n",
|
|
|
324 |
" tokenized_prompt = shard(tokenized_prompt)\n",
|
325 |
"\n",
|
326 |
" # generate images\n",
|
|
|
327 |
" images = []\n",
|
328 |
+
" for i in tqdm(range(num_images // jax.device_count()), desc='Generating Images):\n",
|
329 |
" key, subkey = jax.random.split(key)\n",
|
330 |
" encoded_images = p_generate(tokenized_prompt, shard_prng_key(subkey), model_params)\n",
|
331 |
" encoded_images = encoded_images.sequences[..., 1:]\n",
|
|
|
341 |
" images_per_prompt_indices = np.asarray(range(0, len(images), batch_size))\n",
|
342 |
" clip_inputs['pixel_values'] = jnp.concatenate(list(clip_inputs['pixel_values'][images_per_prompt_indices + i] for i in range(batch_size)))\n",
|
343 |
" clip_inputs = shard(clip_inputs)\n",
|
344 |
+
" logits = p_clip(clip_inputs, clip_params)\n",
|
345 |
" logits = logits.reshape(-1, num_images)\n",
|
346 |
" top_scores = logits.argsort()[:, -top_k:][..., ::-1]\n",
|
347 |
" logits = jax.device_get(logits)\n",
|
|
|
361 |
" images_per_prompt_indices = np.asarray(range(0, len(images), batch_size))\n",
|
362 |
" clip_inputs['pixel_values'] = jnp.concatenate(list(clip_inputs['pixel_values'][images_per_prompt_indices + i] for i in range(batch_size)))\n",
|
363 |
" clip_inputs = shard(clip_inputs)\n",
|
364 |
+
" logits = p_clip32(clip_inputs, clip32_params)\n",
|
365 |
" logits = logits.reshape(-1, num_images)\n",
|
366 |
" top_scores = logits.argsort()[:, -top_k:][..., ::-1]\n",
|
367 |
" logits = jax.device_get(logits)\n",
|