Organize

.gitattributes

@@ -54,3 +54,191 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.webp filter=lfs diff=lfs merge=lfs -text
 java_messages.jsonl filter=lfs diff=lfs merge=lfs -text
 python_messages.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_131636_164545.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2501084_2533993.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2961810_2994719.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1316360_1349269.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2961810_2994719.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_0_32909.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2237812_2270721.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2731447_2764356.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3093446_3126355.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_888543_921452.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1447996_1480905.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_559453_592362.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_691089_723998.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_756907_789816.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1777086_1809995.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2797265_2830174.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_559453_592362.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_855634_888543.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1184724_1217633.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2007449_2040358.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2731447_2764356.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2994719_3027628.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3060537_3093446.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2073267_2106176.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2533993_2566902.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2863083_2895992.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3027628_3060537.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3192173_3225082.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1250542_1283451.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2830174_2863083.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2895992_2928901.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1349269_1382178.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2171994_2204903.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2303630_2336539.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2928901_2961810.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3093446_3126355.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2928901_2961810.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_625271_658180.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1974540_2007449.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2336539_2369448.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1020179_1053088.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1085997_1118906.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1118906_1151815.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1250542_1283451.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2040358_2073267.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2468175_2501084.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_789816_822725.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1349269_1382178.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2994719_3027628.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1777086_1809995.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2270721_2303630.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1053088_1085997.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1809995_1842904.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2435266_2468175.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2665629_2698538.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3225082_3257991.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1579632_1612541.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1612541_1645450.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2501084_2533993.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3159264_3192173.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1316360_1349269.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1480905_1513814.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2106176_2139085.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_65818_98727.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_98727_131636.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1875813_1908722.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_822725_855634.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_954361_987270.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1118906_1151815.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2073267_2106176.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2468175_2501084.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2698538_2731447.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_954361_987270.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1382178_1415087.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1480905_1513814.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2830174_2863083.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_658180_691089.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_888543_921452.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1085997_1118906.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1283451_1316360.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_197454_230363.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_723998_756907.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2106176_2139085.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_921452_954361.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1382178_1415087.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1447996_1480905.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3060537_3093446.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1546723_1579632.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2435266_2468175.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3225082_3257991.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3027628_3060537.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1842904_1875813.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2599811_2632720.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1217633_1250542.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1744177_1777086.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1974540_2007449.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_592362_625271.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1678359_1711268.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2007449_2040358.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1151815_1184724.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1645450_1678359.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3257991_3290900.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_592362_625271.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1546723_1579632.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1579632_1612541.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1941631_1974540.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2204903_2237812.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_230363_263272.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_394908_427817.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1217633_1250542.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1809995_1842904.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2797265_2830174.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_625271_658180.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2533993_2566902.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2764356_2797265.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3192173_3225082.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_756907_789816.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1513814_1546723.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1941631_1974540.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2764356_2797265.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2895992_2928901.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1612541_1645450.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1842904_1875813.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2566902_2599811.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1283451_1316360.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_131636_164545.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2402357_2435266.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2665629_2698538.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_3126355_3159264.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1415087_1447996.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2270721_2303630.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1053088_1085997.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2632720_2665629.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_723998_756907.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_855634_888543.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_263272_296181.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_32909_65818.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_789816_822725.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1020179_1053088.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1875813_1908722.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2369448_2402357.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2698538_2731447.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1513814_1546723.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1711268_1744177.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2599811_2632720.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2632720_2665629.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2237812_2270721.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1151815_1184724.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2402357_2435266.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_460726_493635.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_691089_723998.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1908722_1941631.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_460726_493635.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1184724_1217633.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_987270_1020179.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1645450_1678359.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2171994_2204903.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_296181_329090.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3257991_3290900.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_822725_855634.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_921452_954361.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_987270_1020179.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2336539_2369448.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2566902_2599811.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_526544_559453.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_164545_197454.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2369448_2402357.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_329090_361999.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_427817_460726.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_526544_559453.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1678359_1711268.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1711268_1744177.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1744177_1777086.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2139085_2171994.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2139085_2171994.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2863083_2895992.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_658180_691089.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_493635_526544.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_1908722_1941631.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2040358_2073267.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_2303630_2336539.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3126355_3159264.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_3159264_3192173.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_361999_394908.jsonl filter=lfs diff=lfs merge=lfs -text
+java/tasky_commits_493635_526544.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_1415087_1447996.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_2204903_2237812.jsonl filter=lfs diff=lfs merge=lfs -text
+python/tasky_commits_python_32909_65818.jsonl filter=lfs diff=lfs merge=lfs -text

inference.py

@@ -0,0 +1,138 @@
+# !pip install -q transformers datasets sentencepiece
+import argparse
+import gc
+import json
+import os
+
+import datasets
+import pandas as pd
+import torch
+from tqdm import tqdm
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+
+TOTAL_NUM_FILES_C4_TRAIN = 1024
+
+
+def parse_args():
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument(
+        "--start",
+        type=int,
+        required=True,
+        help="Starting file number to download. Valid values: 0 - 1023",
+    )
+    parser.add_argument(
+        "--end",
+        type=int,
+        required=True,
+        help="Ending file number to download. Valid values: 0 - 1023",
+    )
+    parser.add_argument("--batch_size", type=int, default=64, help="Batch size")
+    parser.add_argument(
+        "--model_name",
+        type=str,
+        default="taskydata/deberta-v3-base_10xp3nirstbbflanseuni_10xc4",
+        help="Model name",
+    )
+    parser.add_argument(
+        "--local_cache_location",
+        type=str,
+        default="c4_download",
+        help="local cache location from where the dataset will be loaded",
+    )
+    parser.add_argument(
+        "--use_local_cache_location",
+        type=bool,
+        default=True,
+        help="Set True if you want to load the dataset from local cache.",
+    )
+    parser.add_argument(
+        "--clear_dataset_cache",
+        type=bool,
+        default=False,
+        help="Set True if you want to delete the dataset files from the cache after inference.",
+    )
+    parser.add_argument(
+        "--release_memory",
+        type=bool,
+        default=True,
+        help="Set True if you want to release the memory of used variables.",
+    )
+
+    args = parser.parse_args()
+    return args
+
+
+def chunks(l, n):
+    for i in range(0, len(l), n):
+        yield l[i : i + n]
+
+
+def batch_tokenize(data, batch_size):
+    batches = list(chunks(data, batch_size))
+    tokenized_batches = []
+    for batch in batches:
+        # max_length will automatically be set to the max length of the model (512 for deberta)
+        tensor = tokenizer(
+            batch,
+            return_tensors="pt",
+            padding="max_length",
+            truncation=True,
+            max_length=512,
+        )
+        tokenized_batches.append(tensor)
+    return tokenized_batches, batches
+
+
+def batch_inference(data, batch_size=32):
+    preds = []
+    tokenized_batches, batches = batch_tokenize(data, batch_size)
+    for i in tqdm(range(len(batches))):
+        with torch.no_grad():
+            logits = model(**tokenized_batches[i].to(device)).logits.cpu()
+        preds.extend(logits)
+    return preds
+
+
+if __name__ == "__main__":
+    args = parse_args()
+
+    tokenizer = AutoTokenizer.from_pretrained(args.model_name)
+    model = AutoModelForSequenceClassification.from_pretrained(args.model_name)
+    device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
+    model.to(device)
+    model.eval()
+
+    path = "python_messages.jsonl"
+    ds = datasets.load_dataset("json", data_files=[path], ignore_verifications=True)["train"]
+    ds = ds[range(args.start, args.end)]
+    df = pd.DataFrame(ds, index=None)
+    texts = df["message"].to_list()
+    commits = df["commit"].to_list()
+    preds = batch_inference(texts, batch_size=args.batch_size)
+
+    assert len(preds) == len(texts)
+
+    # Write two jsonl files:
+    # 1) Probas for all of C4
+    # 2) Probas + texts for samples predicted as tasky
+    tasky_commits_path = f"tasky_commits_python_{args.start}_{args.end}.jsonl"
+
+    with open(tasky_commits_path, "w") as f:
+        for i in range(len(preds)):
+            predicted_class_id = preds[i].argmax().item()
+            pred = model.config.id2label[predicted_class_id]
+            tasky_proba = torch.softmax(preds[i], dim=-1)[-1].item()
+
+            f.write(
+                json.dumps(
+                    {
+                        "commit": commits[i],
+                        "message": texts[i],
+                        "proba": tasky_proba,
+                    }
+                )
+                + "\n"
+            )
+

inference.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+#SBATCH --job-name=inference
+#SBATCH --nodes=1
+#SBATCH --ntasks-per-node=1          # crucial - only 1 task per dist per node!
+#SBATCH --cpus-per-task=8           # number of cores per tasks
+#SBATCH --hint=nomultithread         # we get physical cores not logical
+#SBATCH --gres=gpu:1                 # number of gpus
+#SBATCH --time 12:00:00             # maximum execution time (HH:MM:SS)
+#SBATCH --output=%x-%j.out           # output file name
+#SBATCH --account=ajs@v100
+
+set -x -e
+
+source $six_ALL_CCFRWORK/start-tr13f-6B3-ml-t0
+echo "START TIME: $(date)"
+
+NUM_SAMPLES=3290900
+NUM_SAMPLES_PER_JOB=32909
+#NUM_SAMPLES_PER_JOB=128
+
+TASK_START=$((($SLURM_ARRAY_TASK_ID) * $NUM_SAMPLES_PER_JOB))
+TASK_END=$(($TASK_START + $NUM_SAMPLES_PER_JOB))
+
+echo "Starting task $SLURM_ARRAY_TASK_ID with start $TASK_START and end $TASK_END"
+python inference.py --start $TASK_START --end $TASK_END