hynky HF staff commited on
Commit
e4890d1
1 Parent(s): cf03781
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. Dockerfile +30 -0
  2. README.md +11 -0
  3. analysis/data/metrics_rank_comparison/avg_kendall_tau_a/gen_ranks.json +3 -0
  4. analysis/data/metrics_rank_comparison/avg_kendall_tau_a/mc_ranks.json +3 -0
  5. analysis/data/metrics_rank_comparison/avg_snr/gen_ranks.json +3 -0
  6. analysis/data/metrics_rank_comparison/avg_snr/mc_ranks.json +3 -0
  7. analysis/data/metrics_rank_comparison/avg_spearman/gen_ranks.json +3 -0
  8. analysis/data/metrics_rank_comparison/avg_spearman/mc_ranks.json +3 -0
  9. analysis/data/metrics_rank_comparison/max_n_std/gen_ranks.json +3 -0
  10. analysis/data/metrics_rank_comparison/max_n_std/mc_ranks.json +3 -0
  11. analysis/data/nanotron_models_all.json +3 -0
  12. analysis/data/nanotron_tasks/ar/acva_ara:_average_data.csv +3 -0
  13. analysis/data/nanotron_tasks/ar/acva_ara:_average_stats.csv +3 -0
  14. analysis/data/nanotron_tasks/ar/alfgahafa_mlqa_ara_cf_data.csv +3 -0
  15. analysis/data/nanotron_tasks/ar/alfgahafa_mlqa_ara_cf_stats.csv +3 -0
  16. analysis/data/nanotron_tasks/ar/alghafa_arc_ara_cf:easy_data.csv +3 -0
  17. analysis/data/nanotron_tasks/ar/alghafa_arc_ara_cf:easy_stats.csv +3 -0
  18. analysis/data/nanotron_tasks/ar/alghafa_exams_ara_cf:_average_data.csv +3 -0
  19. analysis/data/nanotron_tasks/ar/alghafa_exams_ara_cf:_average_stats.csv +3 -0
  20. analysis/data/nanotron_tasks/ar/alghafa_facts_ara_cf_data.csv +3 -0
  21. analysis/data/nanotron_tasks/ar/alghafa_facts_ara_cf_stats.csv +3 -0
  22. analysis/data/nanotron_tasks/ar/alghafa_meta_dialects_ara_cf_data.csv +3 -0
  23. analysis/data/nanotron_tasks/ar/alghafa_meta_dialects_ara_cf_stats.csv +3 -0
  24. analysis/data/nanotron_tasks/ar/alghafa_mmlu_ara_cf:_average_data.csv +3 -0
  25. analysis/data/nanotron_tasks/ar/alghafa_mmlu_ara_cf:_average_stats.csv +3 -0
  26. analysis/data/nanotron_tasks/ar/alghafa_openbookqa_ara_cf_data.csv +3 -0
  27. analysis/data/nanotron_tasks/ar/alghafa_openbookqa_ara_cf_stats.csv +3 -0
  28. analysis/data/nanotron_tasks/ar/alghafa_piqa_ara_cf_data.csv +3 -0
  29. analysis/data/nanotron_tasks/ar/alghafa_piqa_ara_cf_stats.csv +3 -0
  30. analysis/data/nanotron_tasks/ar/alghafa_race_ara_cf_data.csv +3 -0
  31. analysis/data/nanotron_tasks/ar/alghafa_race_ara_cf_stats.csv +3 -0
  32. analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_ara_cf_data.csv +3 -0
  33. analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_ara_cf_stats.csv +3 -0
  34. analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_no_neutral_ara_cf_data.csv +3 -0
  35. analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_no_neutral_ara_cf_stats.csv +3 -0
  36. analysis/data/nanotron_tasks/ar/alghafa_sciqa_ara_cf_data.csv +3 -0
  37. analysis/data/nanotron_tasks/ar/alghafa_sciqa_ara_cf_stats.csv +3 -0
  38. analysis/data/nanotron_tasks/ar/alghafa_sentiment_ara_cf_data.csv +3 -0
  39. analysis/data/nanotron_tasks/ar/alghafa_sentiment_ara_cf_stats.csv +3 -0
  40. analysis/data/nanotron_tasks/ar/arcd_ara_data.csv +3 -0
  41. analysis/data/nanotron_tasks/ar/arcd_ara_stats.csv +3 -0
  42. analysis/data/nanotron_tasks/ar/belebele_arb_Arab_cf_data.csv +3 -0
  43. analysis/data/nanotron_tasks/ar/belebele_arb_Arab_cf_stats.csv +3 -0
  44. analysis/data/nanotron_tasks/ar/boolq_ara_data.csv +3 -0
  45. analysis/data/nanotron_tasks/ar/boolq_ara_stats.csv +3 -0
  46. analysis/data/nanotron_tasks/ar/community_arc_hin_cf:challenge_data.csv +3 -0
  47. analysis/data/nanotron_tasks/ar/community_arc_hin_cf:challenge_stats.csv +3 -0
  48. analysis/data/nanotron_tasks/ar/community_arc_hin_cf:easy_data.csv +3 -0
  49. analysis/data/nanotron_tasks/ar/community_arc_hin_cf:easy_stats.csv +3 -0
  50. analysis/data/nanotron_tasks/ar/community_arc_swa_cf:challenge_data.csv +3 -0
Dockerfile ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Use an official Node runtime as the base image for building the application
2
+ FROM node:20 AS build
3
+
4
+ # Set the working directory in the container
5
+ WORKDIR /app
6
+
7
+ # Copy package.json and package-lock.json
8
+ COPY app/package*.json ./
9
+
10
+ # Install dependencies
11
+ RUN npm install
12
+
13
+ # Copy the rest of the application code
14
+ COPY app/ .
15
+ COPY analysis/data ../analysis/data
16
+
17
+ # Build the application
18
+ RUN npm run build
19
+
20
+ # Use an official Nginx runtime as the base image for serving the application
21
+ FROM nginx:alpine
22
+
23
+ # Copy the built application from the build stage
24
+ COPY --from=build /app/dist /usr/share/nginx/html
25
+
26
+ # Expose the port the app runs on
27
+ EXPOSE 80
28
+
29
+ # Command to run the application
30
+ CMD ["nginx", "-g", "daemon off;"]
README.md ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ title: 'FineTasks: Finding signal in a haystack of 200+ multilingual tasks'
3
+ emoji: 📝
4
+ colorFrom: blue
5
+ colorTo: indigo
6
+ sdk: docker
7
+ pinned: false
8
+ header: mini
9
+ app_port: 8080
10
+ thumbnail: https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1/resolve/main/screenshot.jpeg
11
+ ---
analysis/data/metrics_rank_comparison/avg_kendall_tau_a/gen_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:029946adf1e6148f113a7d84c647f064b014a6b40b4ef53051b06bd5c675bdc1
3
+ size 435
analysis/data/metrics_rank_comparison/avg_kendall_tau_a/mc_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25d9071a3c33eabace23029611c4b66e07b8575fec99da1d705c22f4bc6eb75f
3
+ size 1112
analysis/data/metrics_rank_comparison/avg_snr/gen_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7da84530785268472e3b309f484c5fde1d30cedff7363e1125e31bfdcd6a47d2
3
+ size 405
analysis/data/metrics_rank_comparison/avg_snr/mc_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbcecc0bbcc70016c91f01fe0dd02f56276ffde31190214187128924d9a09331
3
+ size 1122
analysis/data/metrics_rank_comparison/avg_spearman/gen_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25315313810b9a3af6463c4e43f0d882391042fb49c8212b3659db39d07a87a8
3
+ size 435
analysis/data/metrics_rank_comparison/avg_spearman/mc_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b102d1ab1fe16264f360d5d128cc49d335d9cf2a19990846b2808355c427885
3
+ size 1065
analysis/data/metrics_rank_comparison/max_n_std/gen_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5ece23a6b8ab6ee28785eb2ce7c4c499dcb9d0110874068c2e371c4e4797c94
3
+ size 435
analysis/data/metrics_rank_comparison/max_n_std/mc_ranks.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e23d4a9ab279bd96b6b6d14c0935a908379f51bdad2b93b2e105feea3795deb4
3
+ size 1126
analysis/data/nanotron_models_all.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba2f6775e5e33f21d7ac9f26444f80e3a0022d2293aa688280755b8fdc600f86
3
+ size 60838684
analysis/data/nanotron_tasks/ar/acva_ara:_average_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee86019600943234de0d00cb7f2cfb5f08adea529e281c47fb11ab39e904fa14
3
+ size 26104
analysis/data/nanotron_tasks/ar/acva_ara:_average_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79551f7eeb2579538604681929741203205e6150f95187ea5319e3e9671f634e
3
+ size 1078
analysis/data/nanotron_tasks/ar/alfgahafa_mlqa_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0753a9fb838808ff6855bfcce87eb7d716d406dff82985e64bd72abf3e0eeed6
3
+ size 20564
analysis/data/nanotron_tasks/ar/alfgahafa_mlqa_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b42cd429953188f1e3a2f61a3bbcd3aa669421bac407a5f2843b9ad3bc287b9b
3
+ size 903
analysis/data/nanotron_tasks/ar/alghafa_arc_ara_cf:easy_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:810a68eb754b4f0a3acae2a34c311676c78d926fd88e34e1c0bb9be949e3aa20
3
+ size 18155
analysis/data/nanotron_tasks/ar/alghafa_arc_ara_cf:easy_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:642e379750d340963d86ff023426787891d7cb494bf135c33be48c0c9897519f
3
+ size 908
analysis/data/nanotron_tasks/ar/alghafa_exams_ara_cf:_average_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf7f6df15ba9e2c552c721bed4d292cf75a8bf6b3f3cbd5f65c9903b99e463d0
3
+ size 24386
analysis/data/nanotron_tasks/ar/alghafa_exams_ara_cf:_average_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b25491622030909b5b075cd7744fcad61fabe7103253c14355710762cbdc6d6
3
+ size 928
analysis/data/nanotron_tasks/ar/alghafa_facts_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3b1f369ae9a64e27702437a049456d90fff09c62133a0232cd146a19bfb1bba
3
+ size 17318
analysis/data/nanotron_tasks/ar/alghafa_facts_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19159aa9195f4a9eebbc6a91431d65dce630ec97edf39e32fe3c0f8dc302e546
3
+ size 834
analysis/data/nanotron_tasks/ar/alghafa_meta_dialects_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:714e3326ff11bfebb268366315fa4b5cf305b9cb8174c451db33773f5ac88d78
3
+ size 18138
analysis/data/nanotron_tasks/ar/alghafa_meta_dialects_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7a37090de73b4fc41f1011e031ad56ff95e3883662275daf8c67656e166b5f9
3
+ size 935
analysis/data/nanotron_tasks/ar/alghafa_mmlu_ara_cf:_average_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6959492567e052a2f9251d092f449dda7ed9118daca1441f5c146e6d2761e10c
3
+ size 23032
analysis/data/nanotron_tasks/ar/alghafa_mmlu_ara_cf:_average_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f579eac42e69687634fa0d27ca738a80d5fd854b4dc33ef069210def32a7394
3
+ size 937
analysis/data/nanotron_tasks/ar/alghafa_openbookqa_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fe35704efef4b670a7f3ec6f64d64aa2e14f387caaa12b7c5da0eda18c4078a
3
+ size 22998
analysis/data/nanotron_tasks/ar/alghafa_openbookqa_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e137221596d32ecd3f77c2f451c87e8ff8743a286816f974e7457290a2cfaec
3
+ size 925
analysis/data/nanotron_tasks/ar/alghafa_piqa_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a436d5d6e494aa37f2f9c4e4f14c2376d97d1c48ed116ca4f9c6f65caf0fbc3f
3
+ size 18478
analysis/data/nanotron_tasks/ar/alghafa_piqa_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f01a72195ce586f94d8dad8210e5accd5459bf83712ad968b149701dfe4b9e8
3
+ size 880
analysis/data/nanotron_tasks/ar/alghafa_race_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80b15aea8264a8bdac278cf2ac0d07cbcc0e8c7ccac150ace4e26dd65471e6fc
3
+ size 18432
analysis/data/nanotron_tasks/ar/alghafa_race_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80ae78c9142cf89d1be55977578c2ca041838a5dbaa736b40958a13730e46ae9
3
+ size 893
analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bdc28b21863e88fcc8fac6245d25da05db90d7da234708b341636400b2584769
3
+ size 18023
analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3291cd3da2406bdcc358764cf42ae1a21cf9d1c07b4daf5998e7695a09317c37
3
+ size 936
analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_no_neutral_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab37a3551307dadb3260bcd768b554954b7451493ce1e07bbaf8465d25f3e09b
3
+ size 16661
analysis/data/nanotron_tasks/ar/alghafa_rating_sentiment_no_neutral_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20a08683aa5cdd6a8a64efcffcf942a4a26dcf2550a2d770b0ff46b8c40ecffe
3
+ size 970
analysis/data/nanotron_tasks/ar/alghafa_sciqa_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a29f2dcacf4f8d53096dddfbeb1f5cb8eb8e8a5354dc2f979a1908c5541ebcd6
3
+ size 23819
analysis/data/nanotron_tasks/ar/alghafa_sciqa_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6cac24425904dec66f4bb38aa34d65d0c1a6bc539baf5ea5300c5f7bc362626
3
+ size 894
analysis/data/nanotron_tasks/ar/alghafa_sentiment_ara_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c69085a0ca2df0adc4f9ec3c3b9857adad82d6749f41d88e3d43ba16e6d936d3
3
+ size 17942
analysis/data/nanotron_tasks/ar/alghafa_sentiment_ara_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1db01058cac603bcb5b0a991b39d9499a3537ab0da1ad36eb1c3b317c8d5ff4
3
+ size 903
analysis/data/nanotron_tasks/ar/arcd_ara_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f5b91b32d5c9a58a34ac035fd9e880de1256f5d0c47edfed7fe591abed789fa
3
+ size 15849
analysis/data/nanotron_tasks/ar/arcd_ara_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d15a351c2e4cf3dcc3372637baafd4821397bb7ab00c81704d91ec8b55e6a31
3
+ size 478
analysis/data/nanotron_tasks/ar/belebele_arb_Arab_cf_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6464f54a9a933b4b47c9c513c907ae358909518998ad5db01d8580578b77a1c6
3
+ size 23912
analysis/data/nanotron_tasks/ar/belebele_arb_Arab_cf_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fda4577c55a3298b9dc6f6a2e148aaf6a4038f8604811ca339f88d3c3f6e7573
3
+ size 903
analysis/data/nanotron_tasks/ar/boolq_ara_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54675b0af158ca756a5c08ea1b6315f757df93827a3a107b4208b135bdf6d8db
3
+ size 18834
analysis/data/nanotron_tasks/ar/boolq_ara_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32457679a2dc38045522dab504c13590c7199bcb452f65acfcf337fbbb3bdc2c
3
+ size 1042
analysis/data/nanotron_tasks/ar/community_arc_hin_cf:challenge_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
3
+ size 14390
analysis/data/nanotron_tasks/ar/community_arc_hin_cf:challenge_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aabcf0d879390556fa664c0fb532afa47580407e37f2552026dbfffab89ebf57
3
+ size 469
analysis/data/nanotron_tasks/ar/community_arc_hin_cf:easy_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
3
+ size 14390
analysis/data/nanotron_tasks/ar/community_arc_hin_cf:easy_stats.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d213b818d04764e3bd9a0d0ad57ab9e0e38fe26d8db51942c12c1b8eb92f3636
3
+ size 449
analysis/data/nanotron_tasks/ar/community_arc_swa_cf:challenge_data.csv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2649f9802e39e19da555d2e42851281cca18826534d23246bffd8b15a43e326a
3
+ size 14390