Title: Models fail to load · Issue #21 · bigcode-project/starcoder.cpp · GitHub
Open Graph Title: Models fail to load · Issue #21 · bigcode-project/starcoder.cpp
X Title: Models fail to load · Issue #21 · bigcode-project/starcoder.cpp
Description: I'm getting the following error in the final step of the quickstart: unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file Input line: ./main -m models/bigcode/gpt_bigcode-santacoder-ggml.bin -p "def fibonnaci(" --top_k 0 --t...
Open Graph Description: I'm getting the following error in the final step of the quickstart: unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file Input line: ./main -m models/bigcode/gpt_bigcode-santacoder-gg...
X Description: I'm getting the following error in the final step of the quickstart: unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file Input line: ./main -m models/bigcode/gpt_bigcode-s...
Opengraph URL: https://github.com/bigcode-project/starcoder.cpp/issues/21
X: @github
Domain: patch-diff.githubusercontent.com
{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Models fail to load","articleBody":"I'm getting the following error in the final step of the quickstart:\r\n\r\n`unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file`\r\n\r\nInput line: \r\n`./main -m models/bigcode/gpt_bigcode-santacoder-ggml.bin -p \"def fibonnaci(\" --top_k 0 --top_p 0.95 --temp 0.2`\r\n\r\nOutput:\r\n```\r\nmain: seed = 1687068338\r\nstarcoder_model_load: loading model from 'models/bigcode/gpt_bigcode-santacoder-ggml-q4_1.bin'\r\nstarcoder_model_load: n_vocab = 49280\r\nstarcoder_model_load: n_ctx = 2048\r\nstarcoder_model_load: n_embd = 2048\r\nstarcoder_model_load: n_head = 16\r\nstarcoder_model_load: n_layer = 24\r\nstarcoder_model_load: ftype = 1003\r\nstarcoder_model_load: qntvr = 1\r\nstarcoder_model_load: ggml ctx size = 1794.97 MB\r\nstarcoder_model_load: memory size = 768.00 MB, n_mem = 49152\r\nstarcoder_model_load: unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file\r\nmain: failed to load model from 'models/bigcode/gpt_bigcode-santacoder-ggml-q4_1.bin'\r\n```\r\n\r\nNotable differences from the sample output:\r\n- `starcoder_model_load: ftype = 1` in my output vs `starcoder_model_load: ftype = 3`\r\n(quanitzed models were produced with `./quantize models/bigcode/gpt_bigcode-santacoder-ggml.bin models/bigcode/gpt_bigcode-santacoder-ggml-q4_1.bin 3`; non-quanitzed model fails with a similar error)\r\n- `starcoder_model_load: qntvr = 1` in my output vs. no info on `qntvr` in the sample output\r\n\r\nOther notes:\r\n- this is running on a 2019 Intel MBP, not an M1\r\n- `conda list` is reproduced below in case I'm somehow missing a dependency\r\n```\r\n# Name Version Build Channel\r\naccelerate 0.20.3 pyhd8ed1ab_0 conda-forge\r\nblas 1.0 mkl\r\nbrotlipy 0.7.0 py310hca72f7f_1002\r\nbzip2 1.0.8 h1de35cc_0\r\nca-certificates 2023.5.7 h8857fd0_0 conda-forge\r\ncertifi 2023.5.7 pyhd8ed1ab_0 conda-forge\r\ncffi 1.15.1 py310h6c40b1e_3\r\ncharset-normalizer 2.0.4 pyhd3eb1b0_0\r\nclick 8.0.4 py310hecd8cb5_0\r\ncryptography 39.0.1 py310hf6deb26_2\r\ndataclasses 0.8 pyh6d0b6a4_7\r\nfilelock 3.9.0 py310hecd8cb5_0\r\nfuture 0.18.3 py310hecd8cb5_0\r\nhuggingface_hub 0.15.1 py_0 huggingface\r\nidna 3.4 py310hecd8cb5_0\r\nimportlib-metadata 6.0.0 py310hecd8cb5_0\r\nimportlib_metadata 6.0.0 hd3eb1b0_0\r\nintel-openmp 2023.1.0 ha357a0b_43547\r\njoblib 1.2.0 py310hecd8cb5_0\r\nlibcxx 14.0.6 h9765a3e_0\r\nlibffi 3.4.4 hecd8cb5_0\r\nlibgfortran 5.0.0 11_3_0_hecd8cb5_28\r\nlibgfortran5 11.3.0 h9dfd629_28\r\nlibopenblas 0.3.21 h54e7dc3_0\r\nlibprotobuf 3.20.3 hfff2838_0\r\nlibuv 1.44.2 h6c40b1e_0\r\nllvm-openmp 14.0.6 h0dcd299_0\r\nmkl 2023.1.0 h59209a4_43558\r\nmkl-service 2.4.0 py310h6c40b1e_1\r\nmkl_fft 1.3.6 py310h3ea8b11_1\r\nmkl_random 1.2.2 py310h3ea8b11_1\r\nncurses 6.4 hcec6c5f_0\r\nninja 1.10.2 hecd8cb5_5\r\nninja-base 1.10.2 haf03e11_5\r\nnumpy 1.24.3 py310h827a554_1\r\nnumpy-base 1.24.3 py310ha186be2_1\r\nopenssl 3.1.1 h8a1eda9_1 conda-forge\r\npackaging 23.0 py310hecd8cb5_0\r\npip 23.1.2 py310hecd8cb5_0\r\nprotobuf 3.20.3 py310hcec6c5f_0\r\npsutil 5.9.5 py310h90acd4f_0 conda-forge\r\npycparser 2.21 pyhd3eb1b0_0\r\npyopenssl 23.0.0 py310hecd8cb5_0\r\npysocks 1.7.1 py310hecd8cb5_0\r\npython 3.10.11 h5ee71fb_3\r\npython_abi 3.10 2_cp310 conda-forge\r\npytorch 1.13.1 cpu_py310h9e40b02_0\r\npyyaml 6.0 py310h6c40b1e_1\r\nreadline 8.2 hca72f7f_0\r\nregex 2022.7.9 py310hca72f7f_0\r\nrequests 2.29.0 py310hecd8cb5_0\r\nsacremoses master py_0 huggingface\r\nsetuptools 67.8.0 py310hecd8cb5_0\r\nsix 1.16.0 pyhd3eb1b0_1\r\nsqlite 3.41.2 h6c40b1e_0\r\ntbb 2021.8.0 ha357a0b_0\r\ntk 8.6.12 h5d9f67b_0\r\ntokenizers 0.11.4 py310h8776b5c_1\r\ntqdm 4.65.0 py310h20db666_0\r\ntransformers 4.28.1 py_0 huggingface\r\ntyping-extensions 4.6.3 py310hecd8cb5_0\r\ntyping_extensions 4.6.3 py310hecd8cb5_0\r\ntzdata 2023c h04d1e81_0\r\nurllib3 1.26.16 py310hecd8cb5_0\r\nwheel 0.38.4 py310hecd8cb5_0\r\nxz 5.4.2 h6c40b1e_0\r\nyaml 0.2.5 haf1e3a3_0\r\nzipp 3.11.0 py310hecd8cb5_0\r\nzlib 1.2.13 h4dc903c_0\r\n```","author":{"url":"https://github.com/peter-ciccolo","@type":"Person","name":"peter-ciccolo"},"datePublished":"2023-06-18T06:08:49.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/21/starcoder.cpp/issues/21"}
| route-pattern | /_view_fragments/issues/show/:user_id/:repository/:id/issue_layout(.:format) |
| route-controller | voltron_issues_fragments |
| route-action | issue_layout |
| fetch-nonce | v2:4617a37a-fae5-9c34-7752-25fea74342a7 |
| current-catalog-service-hash | 81bb79d38c15960b92d99bca9288a9108c7a47b18f2423d0f6438c5b7bcd2114 |
| request-id | BC86:166B1C:35CB09D:4B0078D:69711ED0 |
| html-safe-nonce | 43ba589394bbbeb930502ae9f684f630e988780fc50ce8093aa2f53f10239993 |
| visitor-payload | eyJyZWZlcnJlciI6IiIsInJlcXVlc3RfaWQiOiJCQzg2OjE2NkIxQzozNUNCMDlEOjRCMDA3OEQ6Njk3MTFFRDAiLCJ2aXNpdG9yX2lkIjoiNTk1NDU2NTI4MjUzNjE2OTE2OCIsInJlZ2lvbl9lZGdlIjoiaWFkIiwicmVnaW9uX3JlbmRlciI6ImlhZCJ9 |
| visitor-hmac | c237634ea7b28fdf765a695a76a8dc78e94462d7ef31538cd7bd6a98e7bab80e |
| hovercard-subject-tag | issue:1762132592 |
| github-keyboard-shortcuts | repository,issues,copilot |
| google-site-verification | Apib7-x98H0j5cPqHWwSMm6dNU4GmODRoqxLiDzdx9I |
| octolytics-url | https://collector.github.com/github/collect |
| analytics-location | / |
| fb:app_id | 1401488693436528 |
| apple-itunes-app | app-id=1477376905, app-argument=https://github.com/_view_fragments/issues/show/bigcode-project/starcoder.cpp/21/issue_layout |
| twitter:image | https://opengraph.githubassets.com/c00f4025eb518217c24c45b1c554ca1e74bbe1eed130da7eb41fcd466a0d87ef/bigcode-project/starcoder.cpp/issues/21 |
| twitter:card | summary_large_image |
| og:image | https://opengraph.githubassets.com/c00f4025eb518217c24c45b1c554ca1e74bbe1eed130da7eb41fcd466a0d87ef/bigcode-project/starcoder.cpp/issues/21 |
| og:image:alt | I'm getting the following error in the final step of the quickstart: unknown tensor 'transformer.h.0.attn.q_attn.weight' in model file Input line: ./main -m models/bigcode/gpt_bigcode-santacoder-gg... |
| og:image:width | 1200 |
| og:image:height | 600 |
| og:site_name | GitHub |
| og:type | object |
| og:author:username | peter-ciccolo |
| hostname | github.com |
| expected-hostname | github.com |
| None | 73ca43ac10b0f114fadc2d2cf702bc0c6453174aa5f9dad3a1d131b428870c6d |
| turbo-cache-control | no-preview |
| go-import | github.com/bigcode-project/starcoder.cpp git https://github.com/bigcode-project/starcoder.cpp.git |
| octolytics-dimension-user_id | 110470554 |
| octolytics-dimension-user_login | bigcode-project |
| octolytics-dimension-repository_id | 640843237 |
| octolytics-dimension-repository_nwo | bigcode-project/starcoder.cpp |
| octolytics-dimension-repository_public | true |
| octolytics-dimension-repository_is_fork | false |
| octolytics-dimension-repository_network_root_id | 640843237 |
| octolytics-dimension-repository_network_root_nwo | bigcode-project/starcoder.cpp |
| turbo-body-classes | logged-out env-production page-responsive |
| disable-turbo | false |
| browser-stats-url | https://api.github.com/_private/browser/stats |
| browser-errors-url | https://api.github.com/_private/browser/errors |
| release | 8c16e50bde10edd19f5772e9b3804df6d7bf9bfe |
| ui-target | full |
| theme-color | #1e2327 |
| color-scheme | light dark |
Links:
Viewport: width=device-width