Add code eval dataset and metric

[rishab-partha]

trafficstars

What does this PR do?

Provide ICL datasets and metrics for code evaluation, along with unit tests.

What issue(s) does this change relate to?

CO-2160

[rishab-partha]

Manual test: `****************************** Config: node_name: a100-40sxm-h15-02 num_gpus_per_node: 4 num_nodes: 1 rank_zero_seed: 2021406292

---

[Eval batch=1/11] Eval on humaneval/0-shot data [Eval batch=2/11] Eval on humaneval/0-shot data [Eval batch=3/11] Eval on humaneval/0-shot data [Eval batch=4/11] Eval on humaneval/0-shot data [Eval batch=5/11] Eval on humaneval/0-shot data [Eval batch=6/11] Eval on humaneval/0-shot data [Eval batch=7/11] Eval on humaneval/0-shot data [Eval batch=8/11] Eval on humaneval/0-shot data [Eval batch=9/11] Eval on humaneval/0-shot data [Eval batch=10/11] Eval on humaneval/0-shot data /mnt/workdisk/rishab/composer/composer/core/data_spec.py:35: UserWarning: Cannot split tensor of length 1 into batches of size 4. As it is smaller, no splitting will be done. This may happen on the last batch of a dataset if it is a smaller size than the microbatch size. warnings.warn(f'Cannot split tensor of length {len(t)} into batches of size {microbatch_size}. ' /mnt/workdisk/rishab/composer/composer/core/data_spec.py:26: UserWarning: Cannot split list of length 1 into batches of size 4. As it is smaller, no splitting will be done. This may happen on the last batch of a dataset if it is a smaller size than the microbatch size. warnings.warn(f'Cannot split list of length {len(l)} into batches of size {microbatch_size}. ' [Eval batch=11/11] Eval on humaneval/0-shot data: Eval metrics/humaneval/0-shot/InContextLearningCodeEvalAccuracy: 0.1159 Ran eval in: 846.955498456955 seconds metrics/humaneval/0-shot/InContextLearningCodeEvalAccuracy: 0.11585365980863571`