Anurag1734/cuda-error-resolution-analysis
07
1[2 {3 "post_stream": {4 "posts": [5 {6 "id": 208695,7 "name": "Dylan",8 "username": "Dylan_Yung",9 "avatar_template": "/user_avatar/discuss.pytorch.org/dylan_yung/{size}/15225_2.png",10 "created_at": "2020-07-01T19:32:03.827Z",11 "cooked": "<pre><code class=\"lang-auto\">def sample_model(x: list):\n log_prob_sum = 0\n padded_len = len(max(x, key=len))\n padded_x = list(map(lambda s: [char for char in s] + [PAD] * (padded_len - len(s)), x))\n src = indexTensor(padded_x, padded_len, CHARACTERS).to(DEVICE)\n lng = lengthTensor(x).to(DEVICE)\n hidden = encoder.forward(src, lng)\n lstm_input = targetTensor([SOS] * MINI_BATCH_SZ, 1, CHARACTERS).to(DEVICE)\n names = [''] * MINI_BATCH_SZ\n\n # padded_len + 1 since as length of word increases the Levenshtein distance size goes down\n for i in range(padded_len + 1):\n lstm_probs, hidden = decoder.forward(lstm_input, hidden)\n categorical = torch.distributions.Categorical(\n probs=lstm_probs.squeeze().exp())\n sample = categorical.sample()\n log_prob_sum += categorical.log_prob(sample).sum()\n for j in range(MINI_BATCH_SZ):\n names[j] += CHARACTERS[sample[j].item()]\n lstm_input = sample.unsqueeze(0)\n\n return names, log_prob_sum\n\ndef iterate_train(dl: DataLoader, path: str = \"Checkpoints/\"):\n all_losses = []\n num_model_iterations = 0\n scores_list = []\n\n for epoch_index in range(1, ITER + 1):\n for batch_index, x in enumerate(dl):\n # Zero gradient in models\n encoder_opt.zero_grad()\n decoder_opt.zero_grad()\n\n # Generate noised outputs\n generated_names, log_prob_sum = sample_model(x)\n\n # Split generated names\n noised_list = [name.split(EOS)[0] for name in generated_names]\n\n # Get summary stats of batch\n sample_stats_sum_tensor = get_summary_stats_tensor(noised_list, x)\n\n # Score batch\n distance = torch.dist(sample_stats_sum_tensor, obs_stats_sum_tensor, p=2).detach()\n score = distance * log_prob_sum\n scores_list.append(score)\n\n if batch_index % NUM_SAMPLE == 0:\n # Multiply be -1 because doing gradient descent \n reinforce_loss = -1 * torch.mean(torch.FloatTensor(scores_list))\n reinforce_loss.backward()\n\n encoder_opt.step()\n decoder_opt.step()\n\n\n # Zero out metrics\n scores_list = []\n</code></pre>\n<p>I’m getting an error at the backward. I’m assuming it’s cause I’m multiplying a detached value by a non-detached one? But distance needs to be detached cause the way it’s calculated doesn’t allow a gradient to flow through it. But log_prob_sum does have a gradient flowing through it.</p>",12 "post_number": 1,13 "post_type": 1,14 "posts_count": 3,15 "updated_at": "2020-07-01T19:32:03.827Z",16 "reply_count": 1,17 "reply_to_post_number": null,18 "quote_count": 0,19 "incoming_link_count": 17,20 "reads": 8,21 "readers_count": 7,22 "score": 91.6,23 "yours": false,24 "topic_id": 87638,25 "topic_slug": "reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn",26 "display_username": "Dylan",27 "primary_group_name": null,28 "flair_name": null,29 "flair_url": null,30 "flair_bg_color": null,31 "flair_color": null,32 "flair_group_id": null,33 "badges_granted": [],34 "version": 1,35 "can_edit": false,36 "can_delete": false,37 "can_recover": false,38 "can_see_hidden_post": false,39 "can_wiki": false,40 "read": true,41 "user_title": "",42 "bookmarked": false,43 "actions_summary": [],44 "moderator": false,45 "admin": false,46 "staff": false,47 "user_id": 25820,48 "hidden": false,49 "trust_level": 1,50 "deleted_at": null,51 "user_deleted": false,52 "edit_reason": null,53 "can_view_edit_history": true,54 "wiki": false,55 "post_url": "/t/reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn/87638/1",56 "can_accept_answer": false,57 "can_unaccept_answer": false,58 "accepted_answer": false,59 "topic_accepted_answer": true,60 "can_vote": false61 },62 {63 "id": 208718,64 "name": "Alex",65 "username": "googlebot",66 "avatar_template": "/letter_avatar_proxy/v4/letter/g/ba8739/{size}.png",67 "created_at": "2020-07-01T20:56:10.896Z",68 "cooked": "<aside class=\"quote no-group\" data-username=\"Dylan_Yung\" data-post=\"1\" data-topic=\"87638\">\n<div class=\"title\">\n<div class=\"quote-controls\"></div>\n<img loading=\"lazy\" alt=\"\" width=\"24\" height=\"24\" src=\"https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/dylan_yung/48/15225_2.png\" class=\"avatar\"> Dylan_Yung:</div>\n<blockquote>\n<p><code>torch.mean(torch.FloatTensor(scores_list))</code></p>\n</blockquote>\n</aside>\n<p>try</p>\n<pre><code class=\"lang-auto\">torch.cat(scores_list).mean()\n</code></pre>",69 "post_number": 2,70 "post_type": 1,71 "posts_count": 3,72 "updated_at": "2020-07-05T06:43:26.253Z",73 "reply_count": 1,74 "reply_to_post_number": null,75 "quote_count": 1,76 "incoming_link_count": 0,77 "reads": 5,78 "readers_count": 4,79 "score": 21.0,80 "yours": false,81 "topic_id": 87638,82 "topic_slug": "reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn",83 "display_username": "Alex",84 "primary_group_name": null,85 "flair_name": null,86 "flair_url": null,87 "flair_bg_color": null,88 "flair_color": null,89 "flair_group_id": null,90 "badges_granted": [],91 "version": 1,92 "can_edit": false,93 "can_delete": false,94 "can_recover": false,95 "can_see_hidden_post": false,96 "can_wiki": false,97 "read": true,98 "user_title": null,99 "bookmarked": false,100 "actions_summary": [101 {102 "id": 2,103 "count": 1104 }105 ],106 "moderator": false,107 "admin": false,108 "staff": false,109 "user_id": 29375,110 "hidden": false,111 "trust_level": 2,112 "deleted_at": null,113 "user_deleted": false,114 "edit_reason": null,115 "can_view_edit_history": true,116 "wiki": false,117 "post_url": "/t/reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn/87638/2",118 "can_accept_answer": false,119 "can_unaccept_answer": false,120 "accepted_answer": true,121 "topic_accepted_answer": true122 },123 {124 "id": 209513,125 "name": "Dylan",126 "username": "Dylan_Yung",127 "avatar_template": "/user_avatar/discuss.pytorch.org/dylan_yung/{size}/15225_2.png",128 "created_at": "2020-07-05T06:43:43.653Z",129 "cooked": "<p>it should actually be torch.stack(), but thanks!</p>",130 "post_number": 4,131 "post_type": 1,132 "posts_count": 3,133 "updated_at": "2020-07-05T06:43:43.653Z",134 "reply_count": 0,135 "reply_to_post_number": 2,136 "quote_count": 0,137 "incoming_link_count": 1,138 "reads": 4,139 "readers_count": 3,140 "score": 5.8,141 "yours": false,142 "topic_id": 87638,143 "topic_slug": "reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn",144 "display_username": "Dylan",145 "primary_group_name": null,146 "flair_name": null,147 "flair_url": null,148 "flair_bg_color": null,149 "flair_color": null,150 "flair_group_id": null,151 "badges_granted": [],152 "version": 1,153 "can_edit": false,154 "can_delete": false,155 "can_recover": false,156 "can_see_hidden_post": false,157 "can_wiki": false,158 "read": true,159 "user_title": "",160 "reply_to_user": {161 "id": 29375,162 "username": "googlebot",163 "name": "Alex",164 "avatar_template": "/letter_avatar_proxy/v4/letter/g/ba8739/{size}.png"165 },166 "bookmarked": false,167 "actions_summary": [],168 "moderator": false,169 "admin": false,170 "staff": false,171 "user_id": 25820,172 "hidden": false,173 "trust_level": 1,174 "deleted_at": null,175 "user_deleted": false,176 "edit_reason": null,177 "can_view_edit_history": true,178 "wiki": false,179 "post_url": "/t/reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn/87638/4",180 "can_accept_answer": false,181 "can_unaccept_answer": false,182 "accepted_answer": false,183 "topic_accepted_answer": true184 }185 ],186 "stream": [187 208695,188 208718,189 209513190 ]191 },192 "timeline_lookup": [193 [194 1,195 1942196 ],197 [198 3,199 1939200 ]201 ],202 "suggested_topics": [203 {204 "fancy_title": "Randomly masking a variable number of selected positions in a 2D tensor",205 "id": 215554,206 "title": "Randomly masking a variable number of selected positions in a 2D tensor",207 "slug": "randomly-masking-a-variable-number-of-selected-positions-in-a-2d-tensor",208 "posts_count": 2,209 "reply_count": 0,210 "highest_post_number": 2,211 "image_url": null,212 "created_at": "2025-01-18T12:00:29.986Z",213 "last_posted_at": "2025-01-19T23:42:00.120Z",214 "bumped": true,215 "bumped_at": "2025-01-19T23:42:00.120Z",216 "archetype": "regular",217 "unseen": false,218 "pinned": false,219 "unpinned": null,220 "visible": true,221 "closed": false,222 "archived": false,223 "bookmarked": null,224 "liked": null,225 "tags_descriptions": {},226 "like_count": 0,227 "views": 50,228 "category_id": 1,229 "featured_link": null,230 "has_accepted_answer": false,231 "posters": [232 {233 "extras": null,234 "description": "Original Poster",235 "user": {236 "id": 53086,237 "username": "Norix",238 "name": "",239 "avatar_template": "/letter_avatar_proxy/v4/letter/n/9f8e36/{size}.png",240 "trust_level": 1241 }242 },243 {244 "extras": "latest",245 "description": "Most Recent Poster",246 "user": {247 "id": 18088,248 "username": "KFrank",249 "name": "K. Frank",250 "avatar_template": "/letter_avatar_proxy/v4/letter/k/ecb155/{size}.png",251 "trust_level": 2252 }253 }254 ]255 },256 {257 "fancy_title": "Problem with fork-like multiprocess Dataloader on Ubuntu",258 "id": 213451,259 "title": "Problem with fork-like multiprocess Dataloader on Ubuntu",260 "slug": "problem-with-fork-like-multiprocess-dataloader-on-ubuntu",261 "posts_count": 1,262 "reply_count": 0,263 "highest_post_number": 1,264 "image_url": null,265 "created_at": "2024-11-26T09:11:42.083Z",266 "last_posted_at": "2024-11-26T09:11:42.224Z",267 "bumped": true,268 "bumped_at": "2024-11-26T09:22:55.705Z",269 "archetype": "regular",270 "unseen": false,271 "pinned": false,272 "unpinned": null,273 "visible": true,274 "closed": false,275 "archived": false,276 "bookmarked": null,277 "liked": null,278 "tags_descriptions": {},279 "like_count": 0,280 "views": 82,281 "category_id": 1,282 "featured_link": null,283 "has_accepted_answer": false,284 "posters": [285 {286 "extras": "latest single",287 "description": "Original Poster, Most Recent Poster",288 "user": {289 "id": 81132,290 "username": "neyronon",291 "name": "Victor Lg",292 "avatar_template": "/user_avatar/discuss.pytorch.org/neyronon/{size}/74205_2.png",293 "trust_level": 1294 }295 }296 ]297 },298 {299 "fancy_title": "Non blocking copy from CPU to GPU",300 "id": 213522,301 "title": "Non blocking copy from CPU to GPU",302 "slug": "non-blocking-copy-from-cpu-to-gpu",303 "posts_count": 2,304 "reply_count": 0,305 "highest_post_number": 2,306 "image_url": null,307 "created_at": "2024-11-27T13:17:50.447Z",308 "last_posted_at": "2024-11-28T09:42:41.192Z",309 "bumped": true,310 "bumped_at": "2024-11-28T09:42:41.192Z",311 "archetype": "regular",312 "unseen": false,313 "pinned": false,314 "unpinned": null,315 "visible": true,316 "closed": false,317 "archived": false,318 "bookmarked": null,319 "liked": null,320 "tags_descriptions": {},321 "like_count": 0,322 "views": 162,323 "category_id": 1,324 "featured_link": null,325 "has_accepted_answer": false,326 "posters": [327 {328 "extras": "latest single",329 "description": "Original Poster, Most Recent Poster",330 "user": {331 "id": 81162,332 "username": "shira",333 "name": "shira",334 "avatar_template": "/user_avatar/discuss.pytorch.org/shira/{size}/74225_2.png",335 "trust_level": 1336 }337 }338 ]339 },340 {341 "fancy_title": "Float16 matmul works while bfloat16 throws CUBLAS_STATUS_NOT_SUPPORTED when calling cublasGemmStridedBatchedEx",342 "id": 212129,343 "title": "Float16 matmul works while bfloat16 throws CUBLAS_STATUS_NOT_SUPPORTED when calling cublasGemmStridedBatchedEx",344 "slug": "float16-matmul-works-while-bfloat16-throws-cublas-status-not-supported-when-calling-cublasgemmstridedbatchedex",345 "posts_count": 4,346 "reply_count": 2,347 "highest_post_number": 4,348 "image_url": null,349 "created_at": "2024-10-26T12:14:12.725Z",350 "last_posted_at": "2024-10-27T14:43:23.181Z",351 "bumped": true,352 "bumped_at": "2024-10-27T14:43:23.181Z",353 "archetype": "regular",354 "unseen": false,355 "pinned": false,356 "unpinned": null,357 "visible": true,358 "closed": false,359 "archived": false,360 "bookmarked": null,361 "liked": null,362 "tags_descriptions": {},363 "like_count": 2,364 "views": 138,365 "category_id": 1,366 "featured_link": null,367 "has_accepted_answer": true,368 "posters": [369 {370 "extras": null,371 "description": "Original Poster",372 "user": {373 "id": 65936,374 "username": "Butanium",375 "name": "Clement Dumas",376 "avatar_template": "/user_avatar/discuss.pytorch.org/butanium/{size}/60264_2.png",377 "trust_level": 1378 }379 },380 {381 "extras": "latest",382 "description": "Most Recent Poster, Accepted Answer",383 "user": {384 "id": 3534,385 "username": "ptrblck",386 "name": "",387 "avatar_template": "/user_avatar/discuss.pytorch.org/ptrblck/{size}/1823_2.png",388 "admin": true,389 "moderator": true,390 "trust_level": 2391 }392 }393 ]394 },395 {396 "fancy_title": "What are the most effective and reliable ways to load minibatches efficiently from HDD for deep learning training?",397 "id": 219992,398 "title": "What are the most effective and reliable ways to load minibatches efficiently from HDD for deep learning training?",399 "slug": "what-are-the-most-effective-and-reliable-ways-to-load-minibatches-efficiently-from-hdd-for-deep-learning-training",400 "posts_count": 1,401 "reply_count": 0,402 "highest_post_number": 1,403 "image_url": null,404 "created_at": "2025-05-13T15:01:21.490Z",405 "last_posted_at": "2025-05-13T15:01:21.541Z",406 "bumped": true,407 "bumped_at": "2025-05-13T15:01:21.541Z",408 "archetype": "regular",409 "unseen": false,410 "pinned": false,411 "unpinned": null,412 "visible": true,413 "closed": false,414 "archived": false,415 "bookmarked": null,416 "liked": null,417 "tags_descriptions": {},418 "like_count": 0,419 "views": 19,420 "category_id": 1,421 "featured_link": null,422 "has_accepted_answer": false,423 "posters": [424 {425 "extras": "latest single",426 "description": "Original Poster, Most Recent Poster",427 "user": {428 "id": 63594,429 "username": "AlanTuring",430 "name": "",431 "avatar_template": "/letter_avatar_proxy/v4/letter/a/ea5d25/{size}.png",432 "trust_level": 0433 }434 }435 ]436 }437 ],438 "tags_descriptions": {},439 "fancy_title": "REINFROCE - Element 0 of tensors does not require grad and does not have a grad_fn",440 "id": 87638,441 "title": "REINFROCE - Element 0 of tensors does not require grad and does not have a grad_fn",442 "posts_count": 3,443 "created_at": "2020-07-01T19:32:03.750Z",444 "views": 362,445 "reply_count": 2,446 "like_count": 1,447 "last_posted_at": "2020-07-05T06:43:43.653Z",448 "visible": true,449 "closed": false,450 "archived": false,451 "has_summary": false,452 "archetype": "regular",453 "slug": "reinfroce-element-0-of-tensors-does-not-require-grad-and-does-not-have-a-grad-fn",454 "category_id": 1,455 "word_count": 311,456 "deleted_at": null,457 "user_id": 25820,458 "featured_link": null,459 "pinned_globally": false,460 "pinned_at": null,461 "pinned_until": null,462 "image_url": null,463 "slow_mode_seconds": 0,464 "draft": null,465 "draft_key": "topic_87638",466 "draft_sequence": null,467 "unpinned": null,468 "pinned": false,469 "current_post_number": 1,470 "highest_post_number": 4,471 "deleted_by": null,472 "actions_summary": [473 {474 "id": 4,475 "count": 0,476 "hidden": false,477 "can_act": false478 },479 {480 "id": 8,481 "count": 0,482 "hidden": false,483 "can_act": false484 },485 {486 "id": 10,487 "count": 0,488 "hidden": false,489 "can_act": false490 },491 {492 "id": 7,493 "count": 0,494 "hidden": false,495 "can_act": false496 }497 ],498 "chunk_size": 20,499 "bookmarked": false,500 "topic_timer": null,501 "message_bus_last_id": 0,502 "participant_count": 2,503 "show_read_indicator": false,504 "thumbnails": null,505 "slow_mode_enabled_until": null,506 "accepted_answer": {507 "post_number": 2,508 "username": "googlebot",509 "name": "Alex",510 "excerpt": "try \ntorch.cat(scores_list).mean()"511 },512 "can_vote": false,513 "vote_count": 0,514 "user_voted": false,515 "discourse_zendesk_plugin_zendesk_id": null,516 "discourse_zendesk_plugin_zendesk_url": "https://your-url.zendesk.com/agent/tickets/",517 "details": {518 "can_edit": false,519 "notification_level": 1,520 "participants": [521 {522 "id": 25820,523 "username": "Dylan_Yung",524 "name": "Dylan",525 "avatar_template": "/user_avatar/discuss.pytorch.org/dylan_yung/{size}/15225_2.png",526 "post_count": 2,527 "primary_group_name": null,528 "flair_name": null,529 "flair_url": null,530 "flair_color": null,531 "flair_bg_color": null,532 "flair_group_id": null,533 "trust_level": 1534 },535 {536 "id": 29375,537 "username": "googlebot",538 "name": "Alex",539 "avatar_template": "/letter_avatar_proxy/v4/letter/g/ba8739/{size}.png",540 "post_count": 1,541 "primary_group_name": null,542 "flair_name": null,543 "flair_url": null,544 "flair_color": null,545 "flair_bg_color": null,546 "flair_group_id": null,547 "trust_level": 2548 }549 ],550 "created_by": {551 "id": 25820,552 "username": "Dylan_Yung",553 "name": "Dylan",554 "avatar_template": "/user_avatar/discuss.pytorch.org/dylan_yung/{size}/15225_2.png"555 },556 "last_poster": {557 "id": 25820,558 "username": "Dylan_Yung",559 "name": "Dylan",560 "avatar_template": "/user_avatar/discuss.pytorch.org/dylan_yung/{size}/15225_2.png"561 }562 },563 "bookmarks": []564 },565 {566 "post_stream": {567 "posts": [568 {569 "id": 209505,570 "name": "Fangwei123456",571 "username": "fangwei123456",572 "avatar_template": "/user_avatar/discuss.pytorch.org/fangwei123456/{size}/12650_2.png",573 "created_at": "2020-07-05T04:02:11.229Z",574 "cooked": "<p>How can I find CUDA codes for functions in torch.nn.functional, such as ‘torch.nn.functional.linear()’?</p>",575 "post_number": 1,576 "post_type": 1,577 "posts_count": 1,578 "updated_at": "2020-07-05T04:02:35.051Z",579 "reply_count": 0,580 "reply_to_post_number": null,581 "quote_count": 0,582 "incoming_link_count": 10,583 "reads": 4,584 "readers_count": 3,585 "score": 50.8,586 "yours": false,587 "topic_id": 87985,588 "topic_slug": "how-can-i-find-cuda-codes-for-functions-in-torch-nn-functional",589 "display_username": "Fangwei123456",590 "primary_group_name": null,591 "flair_name": null,592 "flair_url": null,593 "flair_bg_color": null,594 "flair_color": null,595 "flair_group_id": null,596 "badges_granted": [],597 "version": 1,598 "can_edit": false,599 "can_delete": false,600 "can_recover": false,601 "can_see_hidden_post": false,602 "can_wiki": false,603 "read": true,604 "user_title": null,605 "bookmarked": false,606 "actions_summary": [],607 "moderator": false,608 "admin": false,609 "staff": false,610 "user_id": 17622,611 "hidden": false,612 "trust_level": 1,613 "deleted_at": null,614 "user_deleted": false,615 "edit_reason": null,616 "can_view_edit_history": true,617 "wiki": false,618 "post_url": "/t/how-can-i-find-cuda-codes-for-functions-in-torch-nn-functional/87985/1",619 "can_accept_answer": false,620 "can_unaccept_answer": false,621 "accepted_answer": false,622 "topic_accepted_answer": null,623 "can_vote": false624 }625 ],626 "stream": [627 209505628 ]629 },630 "timeline_lookup": [631 [632 1,633 1939634 ]635 ],636 "suggested_topics": [637 {638 "fancy_title": "Stack a vector and another tensor of size 0 or 2",639 "id": 214887,640 "title": "Stack a vector and another tensor of size 0 or 2",641 "slug": "stack-a-vector-and-another-tensor-of-size-0-or-2",642 "posts_count": 2,643 "reply_count": 0,644 "highest_post_number": 2,645 "image_url": null,646 "created_at": "2025-01-02T13:41:32.767Z",647 "last_posted_at": "2025-01-02T13:48:14.740Z",648 "bumped": true,649 "bumped_at": "2025-01-02T13:59:44.279Z",650 "archetype": "regular",651 "unseen": false,652 "pinned": false,653 "unpinned": null,654 "visible": true,655 "closed": false,656 "archived": false,657 "bookmarked": null,658 "liked": null,659 "tags_descriptions": {},660 "like_count": 1,661 "views": 30,662 "category_id": 1,663 "featured_link": null,664 "has_accepted_answer": false,665 "posters": [666 {667 "extras": "latest single",668 "description": "Original Poster, Most Recent Poster",669 "user": {670 "id": 78029,671 "username": "AviZ",672 "name": "",673 "avatar_template": "/letter_avatar_proxy/v4/letter/a/e0b2c6/{size}.png",674 "trust_level": 1675 }676 }677 ]678 },679 {680 "fancy_title": "How can I save the proper embeddings and weights after training?",681 "id": 212912,682 "title": "How can I save the proper embeddings and weights after training?",683 "slug": "how-can-i-save-the-proper-embeddings-and-weights-after-training",684 "posts_count": 2,685 "reply_count": 0,686 "highest_post_number": 2,687 "image_url": null,688 "created_at": "2024-11-13T08:45:07.754Z",689 "last_posted_at": "2024-11-13T12:37:00.417Z",690 "bumped": true,691 "bumped_at": "2024-11-13T15:11:25.864Z",692 "archetype": "regular",693 "unseen": false,694 "pinned": false,695 "unpinned": null,696 "visible": true,697 "closed": false,698 "archived": false,699 "bookmarked": null,700 "liked": null,701 "tags_descriptions": {},702 "like_count": 0,703 "views": 68,704 "category_id": 1,705 "featured_link": null,706 "has_accepted_answer": false,707 "posters": [708 {709 "extras": "latest single",710 "description": "Original Poster, Most Recent Poster",711 "user": {712 "id": 72736,713 "username": "songsong0425",714 "name": "Songyeon Lee",715 "avatar_template": "/user_avatar/discuss.pytorch.org/songsong0425/{size}/67200_2.png",716 "trust_level": 1717 }718 }719 ]720 },721 {722 "fancy_title": "Slurm: Torch not compiled with CUDA enabled",723 "id": 212928,724 "title": "Slurm: Torch not compiled with CUDA enabled",725 "slug": "slurm-torch-not-compiled-with-cuda-enabled",726 "posts_count": 5,727 "reply_count": 6,728 "highest_post_number": 8,729 "image_url": null,730 "created_at": "2024-11-13T13:04:57.351Z",731 "last_posted_at": "2024-11-14T02:51:32.282Z",732 "bumped": true,733 "bumped_at": "2024-11-14T02:51:32.282Z",734 "archetype": "regular",735 "unseen": false,736 "pinned": false,737 "unpinned": null,738 "visible": true,739 "closed": false,740 "archived": false,741 "bookmarked": null,742 "liked": null,743 "tags_descriptions": {},744 "like_count": 0,745 "views": 1298,746 "category_id": 1,747 "featured_link": null,748 "has_accepted_answer": false,749 "posters": [750 {751 "extras": "latest",752 "description": "Original Poster, Most Recent Poster",753 "user": {754 "id": 80887,755 "username": "WuJiayang",756 "name": "",757 "avatar_template": "/letter_avatar_proxy/v4/letter/w/d6d6ee/{size}.png",758 "trust_level": 0759 }760 },761 {762 "extras": null,763 "description": "Frequent Poster",764 "user": {765 "id": 3534,766 "username": "ptrblck",767 "name": "",768 "avatar_template": "/user_avatar/discuss.pytorch.org/ptrblck/{size}/1823_2.png",769 "admin": true,770 "moderator": true,771 "trust_level": 2772 }773 }774 ]775 },776 {777 "fancy_title": "What’s the theoreticl basis of torch.testing tolerance table?",778 "id": 213255,779 "title": "What's the theoreticl basis of torch.testing tolerance table?",780 "slug": "whats-the-theoreticl-basis-of-torch-testing-tolerance-table",781 "posts_count": 1,782 "reply_count": 0,783 "highest_post_number": 1,784 "image_url": "https://discuss.pytorch.org/uploads/default/optimized/3X/7/0/704c9ddf26dc72ffe8f69312a123846fdab78ded_2_1024x736.png",785 "created_at": "2024-11-21T07:42:12.294Z",786 "last_posted_at": "2024-11-21T07:42:12.347Z",787 "bumped": true,788 "bumped_at": "2024-11-21T07:42:12.347Z",789 "archetype": "regular",790 "unseen": false,791 "pinned": false,792 "unpinned": null,793 "visible": true,794 "closed": false,795 "archived": false,796 "bookmarked": null,797 "liked": null,798 "tags_descriptions": {},799 "like_count": 0,800 "views": 25,801 "category_id": 1,802 "featured_link": null,803 "has_accepted_answer": false,804 "posters": [805 {806 "extras": "latest single",807 "description": "Original Poster, Most Recent Poster",808 "user": {809 "id": 81043,810 "username": "Jiexin-Zheng",811 "name": "Jiexin Zheng",812 "avatar_template": "/user_avatar/discuss.pytorch.org/jiexin-zheng/{size}/74108_2.png",813 "trust_level": 0814 }815 }816 ]817 },818 {819 "fancy_title": "How to deal with multiple loss backward when using apply_optimizer_in_backward?",820 "id": 216316,821 "title": "How to deal with multiple loss backward when using apply_optimizer_in_backward?",822 "slug": "how-to-deal-with-multiple-loss-backward-when-using-apply-optimizer-in-backward",823 "posts_count": 3,824 "reply_count": 1,825 "highest_post_number": 3,826 "image_url": null,827 "created_at": "2025-02-06T15:10:19.356Z",828 "last_posted_at": "2025-02-08T17:41:23.987Z",829 "bumped": true,830 "bumped_at": "2025-02-08T17:41:23.987Z",831 "archetype": "regular",832 "unseen": false,833 "pinned": false,834 "unpinned": null,835 "visible": true,836 "closed": false,837 "archived": false,838 "bookmarked": null,839 "liked": null,840 "tags_descriptions": {},841 "like_count": 0,842 "views": 73,843 "category_id": 1,844 "featured_link": null,845 "has_accepted_answer": false,846 "posters": [847 {848 "extras": "latest",849 "description": "Original Poster, Most Recent Poster",850 "user": {851 "id": 12261,852 "username": "biggerfish",853 "name": "",854 "avatar_template": "/letter_avatar_proxy/v4/letter/b/ecae2f/{size}.png",855 "trust_level": 1856 }857 },858 {859 "extras": null,860 "description": "Frequent Poster",861 "user": {862 "id": 41396,863 "username": "soulitzer",864 "name": "",865 "avatar_template": "/letter_avatar_proxy/v4/letter/s/839c29/{size}.png",866 "trust_level": 2867 }868 }869 ]870 }871 ],872 "tags_descriptions": {},873 "fancy_title": "How can I find CUDA codes for functions in torch.nn.functional?",874 "id": 87985,875 "title": "How can I find CUDA codes for functions in torch.nn.functional?",876 "posts_count": 1,877 "created_at": "2020-07-05T04:02:11.180Z",878 "views": 316,879 "reply_count": 0,880 "like_count": 0,881 "last_posted_at": "2020-07-05T04:02:11.229Z",882 "visible": true,883 "closed": false,884 "archived": false,885 "has_summary": false,886 "archetype": "regular",887 "slug": "how-can-i-find-cuda-codes-for-functions-in-torch-nn-functional",888 "category_id": 1,889 "word_count": 18,890 "deleted_at": null,891 "user_id": 17622,892 "featured_link": null,893 "pinned_globally": false,894 "pinned_at": null,895 "pinned_until": null,896 "image_url": null,897 "slow_mode_seconds": 0,898 "draft": null,899 "draft_key": "topic_87985",900 "draft_sequence": null,901 "unpinned": null,902 "pinned": false,903 "current_post_number": 1,904 "highest_post_number": 1,905 "deleted_by": null,906 "actions_summary": [907 {908 "id": 4,909 "count": 0,910 "hidden": false,911 "can_act": false912 },913 {914 "id": 8,915 "count": 0,916 "hidden": false,917 "can_act": false918 },919 {920 "id": 10,921 "count": 0,922 "hidden": false,923 "can_act": false924 },925 {926 "id": 7,927 "count": 0,928 "hidden": false,929 "can_act": false930 }931 ],932 "chunk_size": 20,933 "bookmarked": false,934 "topic_timer": null,935 "message_bus_last_id": 0,936 "participant_count": 1,937 "show_read_indicator": false,938 "thumbnails": null,939 "slow_mode_enabled_until": null,940 "can_vote": false,941 "vote_count": 0,942 "user_voted": false,943 "discourse_zendesk_plugin_zendesk_id": null,944 "discourse_zendesk_plugin_zendesk_url": "https://your-url.zendesk.com/agent/tickets/",945 "details": {946 "can_edit": false,947 "notification_level": 1,948 "participants": [949 {950 "id": 17622,951 "username": "fangwei123456",952 "name": "Fangwei123456",953 "avatar_template": "/user_avatar/discuss.pytorch.org/fangwei123456/{size}/12650_2.png",954 "post_count": 1,955 "primary_group_name": null,956 "flair_name": null,957 "flair_url": null,958 "flair_color": null,959 "flair_bg_color": null,960 "flair_group_id": null,961 "trust_level": 1962 }963 ],964 "created_by": {965 "id": 17622,966 "username": "fangwei123456",967 "name": "Fangwei123456",968 "avatar_template": "/user_avatar/discuss.pytorch.org/fangwei123456/{size}/12650_2.png"969 },970 "last_poster": {971 "id": 17622,972 "username": "fangwei123456",973 "name": "Fangwei123456",974 "avatar_template": "/user_avatar/discuss.pytorch.org/fangwei123456/{size}/12650_2.png"975 }976 },977 "bookmarks": []978 },979 {980 "post_stream": {981 "posts": [982 {983 "id": 61944,984 "name": "tjl",985 "username": "eternity",986 "avatar_template": "/letter_avatar_proxy/v4/letter/e/0ea827/{size}.png",987 "created_at": "2018-08-21T11:01:43.892Z",988 "cooked": "<p>Now,I want to calculate the mean and std of my dataset(about 10000 images).<br>\nI know I can get the mean by:<br>\n<div class=\"lightbox-wrapper\"><a class=\"lightbox\" href=\"https://discuss.pytorch.org/uploads/default/original/2X/7/7ee0de14beccee20b46d1c5a5a0be879116550c9.png\" data-download-href=\"https://discuss.pytorch.org/uploads/default/7ee0de14beccee20b46d1c5a5a0be879116550c9\" title=\"G(%605YC6BLH%24%EF%BF%BDE(DA)2RJK\"><img src=\"https://discuss.pytorch.org/uploads/default/original/2X/7/7ee0de14beccee20b46d1c5a5a0be879116550c9.png\" alt=\"G(%605YC6BLH%24%EF%BF%BDE(DA)2RJK\" data-base62-sha1=\"i6pW8rTpe4xdQznCK28IyaZdBTX\" width=\"443\" height=\"500\" data-dominant-color=\"2F333B\"><div class=\"meta\"><svg class=\"fa d-icon d-icon-far-image svg-icon\" aria-hidden=\"true\"><use href=\"#far-image\"></use></svg><span class=\"filename\">G(%605YC6BLH%24%EF%BF%BDE(DA)2RJK</span><span class=\"informations\">597×673 25.5 KB</span><svg class=\"fa d-icon d-icon-discourse-expand svg-icon\" aria-hidden=\"true\"><use href=\"#discourse-expand\"></use></svg></div></a></div></p>\n<p>However, I do not know how to get <strong>the std</strong> further effectively.<br>\nCan someone tell me how to get it? Better with the corresponding code!<br>\nThanks in advance!</p>",989 "post_number": 1,990 "post_type": 1,991 "posts_count": 4,992 "updated_at": "2018-08-21T11:01:43.892Z",993 "reply_count": 0,994 "reply_to_post_number": null,995 "quote_count": 0,996 "incoming_link_count": 19662,997 "reads": 756,998 "readers_count": 755,999 "score": 98476.2,1000 "yours": false,1001 "topic_id": 23584,1002 "topic_slug": "how-to-calculate-the-mean-and-std-of-my-own-dataset",1003 "display_username": "tjl",1004 "primary_group_name": null,1005 "flair_name": null,1006 "flair_url": null,1007 "flair_bg_color": null,1008 "flair_color": null,1009 "flair_group_id": null,1010 "badges_granted": [],1011 "version": 1,1012 "can_edit": false,1013 "can_delete": false,1014 "can_recover": false,1015 "can_see_hidden_post": false,1016 "can_wiki": false,1017 "link_counts": [1018 {1019 "url": "https://discuss.pytorch.org/uploads/default/original/2X/7/7ee0de14beccee20b46d1c5a5a0be879116550c9.png",1020 "internal": true,1021 "reflection": false,1022 "clicks": 01023 }1024 ],1025 "read": true,1026 "user_title": null,1027 "bookmarked": false,1028 "actions_summary": [1029 {1030 "id": 2,1031 "count": 21032 }1033 ],1034 "moderator": false,1035 "admin": false,1036 "staff": false,1037 "user_id": 10463,1038 "hidden": false,1039 "trust_level": 1,1040 "deleted_at": null,1041 "user_deleted": false,1042 "edit_reason": null,1043 "can_view_edit_history": true,1044 "wiki": false,1045 "post_url": "/t/how-to-calculate-the-mean-and-std-of-my-own-dataset/23584/1",1046 "can_accept_answer": false,1047 "can_unaccept_answer": false,1048 "accepted_answer": false,1049 "topic_accepted_answer": null,1050 "can_vote": false1051 },1052 {1053 "id": 61946,1054 "name": "",1055 "username": "ptrblck",1056 "avatar_template": "/user_avatar/discuss.pytorch.org/ptrblck/{size}/1823_2.png",1057 "created_at": "2018-08-21T11:06:12.883Z",1058 "cooked": "<p>Would <a href=\"https://discuss.pytorch.org/t/about-normalization-using-pre-trained-vgg16-networks/23560/6?u=ptrblck\">this answer</a> from the other thread work?</p>",1059 "post_number": 2,1060 "post_type": 1,1061 "posts_count": 4,1062 "updated_at": "2018-08-21T11:06:12.883Z",1063 "reply_count": 0,1064 "reply_to_post_number": null,1065 "quote_count": 0,1066 "incoming_link_count": 67,1067 "reads": 731,1068 "readers_count": 730,1069 "score": 511.2,1070 "yours": false,1071 "topic_id": 23584,1072 "topic_slug": "how-to-calculate-the-mean-and-std-of-my-own-dataset",1073 "display_username": "",1074 "primary_group_name": null,1075 "flair_name": null,1076 "flair_url": null,1077 "flair_bg_color": null,1078 "flair_color": null,1079 "flair_group_id": null,1080 "badges_granted": [],1081 "version": 1,1082 "can_edit": false,1083 "can_delete": false,1084 "can_recover": false,1085 "can_see_hidden_post": false,1086 "can_wiki": false,1087 "link_counts": [1088 {1089 "url": "https://discuss.pytorch.org/t/about-normalization-using-pre-trained-vgg16-networks/23560/6",1090 "internal": true,1091 "reflection": false,1092 "title": "About Normalization using pre-trained vgg16 networks",1093 "clicks": 83201094 }1095 ],1096 "read": true,1097 "user_title": "",1098 "bookmarked": false,1099 "actions_summary": [1100 {1101 "id": 2,1102 "count": 21103 }1104 ],1105 "moderator": true,1106 "admin": true,1107 "staff": true,1108 "user_id": 3534,1109 "hidden": false,1110 "trust_level": 2,1111 "deleted_at": null,1112 "user_deleted": false,1113 "edit_reason": null,1114 "can_view_edit_history": true,1115 "wiki": false,1116 "post_url": "/t/how-to-calculate-the-mean-and-std-of-my-own-dataset/23584/2",1117 "can_accept_answer": false,1118 "can_unaccept_answer": false,1119 "accepted_answer": false,1120 "topic_accepted_answer": null1121 },1122 {1123 "id": 193091,1124 "name": "sharkdeng",1125 "username": "sharkdeng",1126 "avatar_template": "/user_avatar/discuss.pytorch.org/sharkdeng/{size}/23021_2.png",1127 "created_at": "2020-05-15T09:18:59.542Z",1128 "cooked": "<p>Got same question here. Previously I was using ImageNet fixed Normalize technique. What is the difference between Imagenet and self dataset?<br>\nWhen I used ImageNet Normalize, the image becomes dark(pixels are below 0). I will try to get specific mean and std of my dataset to see if the image would not be black after Normalization.</p>",1129 "post_number": 3,1130 "post_type": 1,1131 "posts_count": 4,1132 "updated_at": "2020-05-15T09:18:59.542Z",1133 "reply_count": 0,1134 "reply_to_post_number": null,1135 "quote_count": 0,1136 "incoming_link_count": 82,1137 "reads": 388,1138 "readers_count": 387,1139 "score": 487.6,1140 "yours": false,1141 "topic_id": 23584,1142 "topic_slug": "how-to-calculate-the-mean-and-std-of-my-own-dataset",1143 "display_username": "sharkdeng",1144 "primary_group_name": null,1145 "flair_name": null,1146 "flair_url": null,1147 "flair_bg_color": null,1148 "flair_color": null,1149 "flair_group_id": null,1150 "badges_granted": [],1151 "version": 1,1152 "can_edit": false,1153 "can_delete": false,1154 "can_recover": false,1155 "can_see_hidden_post": false,1156 "can_wiki": false,1157 "read": true,1158 "user_title": "",1159 "bookmarked": false,1160 "actions_summary": [],1161 "moderator": false,1162 "admin": false,1163 "staff": false,1164 "user_id": 21933,1165 "hidden": false,1166 "trust_level": 2,1167 "deleted_at": null,1168 "user_deleted": false,1169 "edit_reason": null,1170 "can_view_edit_history": true,1171 "wiki": false,1172 "post_url": "/t/how-to-calculate-the-mean-and-std-of-my-own-dataset/23584/3",1173 "can_accept_answer": false,1174 "can_unaccept_answer": false,1175 "accepted_answer": false,1176 "topic_accepted_answer": null1177 },1178 {1179 "id": 209503,1180 "name": "Gorkem Polat",1181 "username": "GorkemP",1182 "avatar_template": "/letter_avatar_proxy/v4/letter/g/a9adbd/{size}.png",1183 "created_at": "2020-07-05T03:36:53.606Z",1184 "cooked": "<p>Just as you did for mean, you can easily adapt your code to calculate standard deviation (after you calculated the means). In addition, if you count the number of pixels (width, height) in the loop, even if your images have different sizes you can get the exact number to divide the sum:</p>\n<pre><code class=\"lang-auto\">R_channel = 0\nG_channel = 0\nB_channel = 0\n\ntotal_pixel = 0\nfor idx in xrange(len(pathDir)):\n filename = pathDir[idx]\n img = imread(os.path.join(filepath, filename))\n\n total_pixel = total_pixel + img.shape[0] * img.shape[1]\n\n R_total = R_total + np.sum((img[:, :, 0] - R_mean) ** 2)\n G_total = G_total + np.sum((img[:, :, 1] - G_mean) ** 2)\n B_total = B_total + np.sum((img[:, :, 2] - B_mean) ** 2)\n\nR_std = sqrt(R_total / total_count)\nG_std = sqrt(G_total / total_count)\nB_std = sqrt(B_total / total_count)\n</code></pre>",1185 "post_number": 4,1186 "post_type": 1,1187 "posts_count": 4,1188 "updated_at": "2020-07-05T03:45:19.203Z",1189 "reply_count": 0,1190 "reply_to_post_number": null,1191 "quote_count": 0,1192 "incoming_link_count": 194,1193 "reads": 304,1194 "readers_count": 303,1195 "score": 1060.8,1196 "yours": false,1197 "topic_id": 23584,1198 "topic_slug": "how-to-calculate-the-mean-and-std-of-my-own-dataset",1199 "display_username": "Gorkem Polat",1200 "primary_group_name": null,