salahh297 commited on
Commit
8be9ddd
·
0 Parent(s):

Super-squash branch 'main' using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ language:
4
+ - ar
5
+ tags:
6
+ - text-classification
7
+ - arabic
8
+ - ensemble
9
+ - doha-ai-challenge
10
+ ---
11
+
12
+ # Doha AI Challenge — fine-tuned checkpoints (Team Salahh)
13
+
14
+ This repository holds the **fine-tuned model checkpoints** for our winning solution to the
15
+ **Google Cloud Doha AI Challenge** (6-target Arabic-news classification, weighted-F1).
16
+
17
+ The solution is an ensemble of Arabic transformer classifiers built on **9 pretrained backbones**,
18
+ each fine-tuned under several configurations/seeds. Every run is trained with 5-fold cross-validation,
19
+ so each of the 42 runs contributes **5 fold checkpoints**:
20
+
21
+ ```
22
+ <tag>/fold0.pt fold1.pt fold2.pt fold3.pt fold4.pt # 42 tags x 5 folds = 210 checkpoints
23
+ ```
24
+
25
+ Each `fold*.pt` is a `state_dict` (bf16) of a multi-task model: a shared encoder + 6 linear
26
+ classification heads (one per target). ~92 GB total.
27
+
28
+ These checkpoints are consumed by the **`solution/`** code of the competition submission package
29
+ (`infer.py` / `reproduce_from_checkpoints.ipynb`), together with `recipe.json` (per-target model sets)
30
+ and `weights.json` (ensemble weights), to reproduce the final `clean.csv`.
31
+
32
+ ---
33
+
34
+ ## 1. Download the checkpoints
35
+
36
+ ```bash
37
+ pip install -U "huggingface_hub[cli]"
38
+
39
+ # into the solution/ folder of the submission package, as checkpoints/<tag>/fold*.pt
40
+ huggingface-cli download salahh297/doha-checkpoints \
41
+ --repo-type model --local-dir solution/checkpoints
42
+ ```
43
+ (Or in Python:)
44
+ ```python
45
+ from huggingface_hub import snapshot_download
46
+ snapshot_download("salahh297/doha-checkpoints", repo_type="model",
47
+ local_dir="solution/checkpoints")
48
+ ```
49
+
50
+ After download, the layout is `solution/checkpoints/<tag>/fold{0..4}.pt`.
51
+
52
+ ---
53
+
54
+ ## 2. Reproduce `clean.csv` from the checkpoints (GPU)
55
+
56
+ You need the competition **`solution/`** folder (code + `recipe.json` + `weights.json` +
57
+ `data/Train.csv`, `data/Test.csv`) alongside the downloaded `checkpoints/`.
58
+
59
+ ```bash
60
+ cd solution
61
+ pip install torch "transformers==4.57.1" accelerate sentencepiece arabert farasapy scikit-learn pandas numpy
62
+
63
+ # script:
64
+ python3 infer.py
65
+ # or the notebook (same result):
66
+ # jupyter nbconvert --to notebook --execute reproduce_from_checkpoints.ipynb
67
+ ```
68
+
69
+ `infer.py` rebuilds each model, loads its five fold checkpoints from `checkpoints/<tag>/fold*.pt`,
70
+ runs them forward on the test set (GPU), averages the five test-probability sets per model, and
71
+ combines the per-target model sets (`recipe.json`) with the ensemble weights (`weights.json`) to
72
+ write `clean.csv`.
73
+
74
+ - **GPU:** A100 40GB, or A100 / H100 80GB (H100 80GB recommended).
75
+ - The farasa-segmented run preprocesses the test text with the farasa segmenter and is slow
76
+ (tens of minutes); all other runs are fast.
77
+ - Fixed seeds are used throughout, so the pipeline is reproducible.
78
+
79
+ ---
80
+
81
+ _Team: Salahh._
arabertL/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b930d0d7af1ddd539f7351a88403106a78d7b924a51c1d61df91533df3386594
3
+ size 739017447
arabertL/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9034272bb154f7272ca52bd6c3019a13d108473785a62762d7d9addb12a04cfa
3
+ size 739017447
arabertL/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f6c4b67c41e0ad795efb024feebc0cca6d2747e86fb2c2335cd10b2e79942f4
3
+ size 739017447
arabertL/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2199e89b9fe1611611553f3aed58d0fdcb010c868ba65bc93299d9dbad6568e1
3
+ size 739017447
arabertL/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eedde3a59a07edb41ffcd1df012ea3ffed5a67a45180a77137e1bf9f1fa368b2
3
+ size 739017447
arb_e8_s2/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1b261aa0f651312ac2b7f67f8e5c9a766868529a45e90653da07005b1791673
3
+ size 325786023
arb_e8_s2/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f500c9737f82b0883df0b522034483ce011cbb584b8ca40d625753368e735d83
3
+ size 325786023
arb_e8_s2/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38dbd34ed1c79e8af6a11ad7b5f22c75d39db8d047c6e061cf4e15c60909973e
3
+ size 325786023
arb_e8_s2/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a87a8dad15416b6d42fca2f9defec30c41f39c4f2749a3ae1a505ec4919b3f0
3
+ size 325786023
arb_e8_s2/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4945ecf666ddf6b0c7ccf2629aa18a9b226a72a055c9e070d18a347cd21b8b3c
3
+ size 325786023
arbertv2/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c537b39fab578a0f116ed8e18d5f4a9a31ded555a85766b351ded5483b5437d7
3
+ size 325786023
arbertv2/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2509f2cba27fc27fd701c9b524e5861428b2ba955d87ec3eb965392f6996f43d
3
+ size 325786023
arbertv2/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5abfa5dfc125dff1ae4ece32c8e9bccad4a6ca22dc7a43fb3f04ba106b1c73e0
3
+ size 325786023
arbertv2/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fb4861a7ff7285255ffa55978baa0f7dcdf0994b2888776638767f5350fccf4
3
+ size 325786023
arbertv2/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2b82f739c8efe46c302485b485d5232b788fb7b112eeb998dae1a068ba9ea14
3
+ size 325786023
arbertv2_ht_e10/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db140e1f2f5ad4feff2f3946a6328645c26fcf422069d946f510c6c70ad7a512
3
+ size 325786023
arbertv2_ht_e10/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe5454b376fcf0909d157781c39d7aa6c5b737e8d3b617ef1cc59bf2f3fb9d02
3
+ size 325786023
arbertv2_ht_e10/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fdf16786c655d68a952bd6f3a30338b8d6aa7fee0ab5eabc4f9facf7a8361b80
3
+ size 325786023
arbertv2_ht_e10/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61a63c6e96f7a1908375d6fbebdb40edc9f1b092d9cbefdcfade4702e092f7d6
3
+ size 325786023
arbertv2_ht_e10/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc2988d8abb8553857df3ef21fe4d0b3287d3348e2c9023e878fac32b90e74f1
3
+ size 325786023
arbertv2_ht_e12/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd9e2b90fb3389368ed980913cb362b2220981c343e05d0698297f9b0c3778ca
3
+ size 325786023
arbertv2_ht_e12/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:126babac443482cd75b0d2f5aa3ddec53cb3f91a1fc67964e1d69422d812d45e
3
+ size 325786023
arbertv2_ht_e12/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:314e31677b9456185ca861c3fe757c4284088841874634e5a3a8b4ef46eda314
3
+ size 325786023
arbertv2_ht_e12/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e3e5bec46ad532e891b4a56da0ba5ce4aa592ded34e1bd0f6df499471eb6757
3
+ size 325786023
arbertv2_ht_e12/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e94028473d5f86897845aef61d49fbb4dcde53a9eb951066db55bea7d630bc86
3
+ size 325786023
arbertv2_ht_e8/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3730352584db365fb435d29e026bce10ad54c5914751f33743e3cffc75d4e360
3
+ size 325786023
arbertv2_ht_e8/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7d00c35f7647cc7e46f44a6a1acbc04e87504fb13b6cb6d7dd0eac447969fc5
3
+ size 325786023
arbertv2_ht_e8/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a04ad532c6b1b2841e40936227133d93369a15952f86af1d1bed6b884f023889
3
+ size 325786023
arbertv2_ht_e8/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc45f0b1e015e36d1792ab480a468403e41b647e6e1008248d4da2bdcd363317
3
+ size 325786023
arbertv2_ht_e8/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b395080760a6f77c21e3ab671f8092f14fe9a8054950754c6c690f8f8b63d9a5
3
+ size 325786023
arbertv2_ht_e8_s3/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5687a7c073ac90723e6cf8620f5d049ddb3001b1c848f89b5f68ac5a0ba0803a
3
+ size 325786023
arbertv2_ht_e8_s3/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee29660999a24646447cba89f2a28f302980267e99164769071a054b7b9ee3dc
3
+ size 325786023
arbertv2_ht_e8_s3/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:356acf4d9ab09511e7de97de6889762430355526c8d6eb098c01061ca8f30b78
3
+ size 325786023
arbertv2_ht_e8_s3/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e775d8c5b0c2ed17f771aeb3feecb4280ad0868f80a7422354dcea62776fc23b
3
+ size 325786023
arbertv2_ht_e8_s3/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d94874ef6840810612e12275d2d529dc863c00249ef06ba170baefce71cdeb0
3
+ size 325786023
arbertv2_ht_e8_s4/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16700d9608cb9516094d6877f0daad55d4f92af1b341871b551f7a043ee9142d
3
+ size 325786023
arbertv2_ht_e8_s4/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c261d6084cbec6328550e80f0a845087fa4d92a578117f0c9a0c6245178b262c
3
+ size 325786023
arbertv2_ht_e8_s4/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c96b403dd91f90166569024d5c1583e1bbd65531126b3592659a6fa7d58eeb2e
3
+ size 325786023
arbertv2_ht_e8_s4/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e8d2abde599385979c544de15eb32e3bfcdac3a66de134771070e856989edf5
3
+ size 325786023
arbertv2_ht_e8_s4/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d238bdf78a3bd38e2e954015475e1fab049706be2d5d7386b5a0ba937de5db71
3
+ size 325786023
base_arabertv02_tapt/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16f4be2545c684d4ce7b5808894c9e9076c8221ae3c5b21978a39b4414c04d1f
3
+ size 270490023
base_arabertv02_tapt/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de361ae6b95ada0b65e8b492d9e71b9339b576ea242b22fbebef1345a1531852
3
+ size 270490023
base_arabertv02_tapt/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0bbaa2e6fafd1923ead1b9b26a93d81f9b3bf19cbee9b2005b600e17d3a15af
3
+ size 270490023
base_arabertv02_tapt/fold3.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f6abadc63fd8d346e3dfab52c935f9d253670dde0fa0db178b9e69837ece00c
3
+ size 270490023
base_arabertv02_tapt/fold4.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38110014ae477b32c976bf6ac18e8f31f0f0294fe360a7539e97a900b9d42e23
3
+ size 270490023
base_e8_s2/fold0.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e1730d08e990f1933aeb90560ea5b8541e5d963affdf21b00c2cf53fa531664
3
+ size 270490023
base_e8_s2/fold1.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8e31a537af7977d4b562b3cce0678414a497a4f10b830afd07603321aef6b04
3
+ size 270490023
base_e8_s2/fold2.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9e51108358c0cc3a621fcc6a822b8e870b0b0311d4ad5ac752019614277d933
3
+ size 270490023