updates readme for releases

[edbeeching]

adds more details of the datasets and contributing.

[qgallouedec]

srun accelerate launch $WORK/gia/scripts/train_jat_tokenized.py \
    --output_dir $WORK/checkpoints/jat_small_v100 \
    --model_name_or_path jat-project/jat-small \
    --tasks all \
    --trust_remote_code \
    --per_device_train_batch_size 20 \
    --gradient_accumulation_steps 2 \
    --save_steps 10000 \
    --run_name train_jat_small \
    --logging_steps 100 \
    --logging_first_step \
    --dispatch_batches False \
    --dataloader_num_workers 16 \
    --max_steps 250000 \
    --resume_from_checkpoint checkpoints/jat_small_v100/checkpoint-220000