Skip to content

Add new_results with OPT-2.7B Java evaluation and fix inference scripts - #1

Open
Hrushik007 wants to merge 1 commit into
Large-scale-Sustainable-Computing-LSC:mainfrom
Hrushik007:dev
Open

Hrushik007 wants to merge 1 commit into
Large-scale-Sustainable-Computing-LSC:mainfrom
Hrushik007:dev

Conversation

@Hrushik007

Copy link
Copy Markdown

No description provided.

Copilot AI lite review requested due to automatic review settings August 18, 2026 04:22

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Adds replicated Java evaluation outputs for OPT-2.7B (fixed-exit) and supporting plotting scripts, while updating dependencies and adjusting the evaluator’s OPT model import.

Changes:

  • Updated Python dependencies (version pin changes and removal of an NVIDIA CUDA wheel pin).
  • Added OPT-2.7B Java fixed-exit evaluation result JSONs (per-exit + aggregated) under both results/ and new_results/.
  • Added Pareto plotting scripts for the Java fixed-exit metrics.

Reviewed changes

Copilot reviewed 28 out of 41 changed files in this pull request and generated 3 comments.

Show a summary per file
File Description
src/requirements.txt Adjusts dependency pins (CodeBLEU / tree-sitter) and removes a CUDA wheel pin.
src/eval/evaluator.py Changes OPT fixed-exit model import (now aliases OPTEESingleHead).
results/replication/opt_java/eval_exit_3.json Adds replication metrics for exit 3.
results/replication/opt_java/eval_exit_5.json Adds replication metrics for exit 5.
results/replication/opt_java/eval_exit_7.json Adds replication metrics for exit 7.
results/replication/opt_java/eval_exit_9.json Adds replication metrics for exit 9.
results/replication/opt_java/eval_exit_11.json Adds replication metrics for exit 11.
results/replication/opt_java/eval_exit_13.json Adds replication metrics for exit 13.
results/replication/opt_java/eval_exit_15.json Adds replication metrics for exit 15.
results/replication/opt_java/eval_exit_19.json Adds replication metrics for exit 19.
results/replication/opt_java/eval_exit_23.json Adds replication metrics for exit 23.
results/replication/opt_java/eval_exit_27.json Adds replication metrics for exit 27.
results/replication/opt_java/eval_exit_31.json Adds replication metrics for exit 31.
results/replication/opt_java/eval_all_fixed_exits.json Adds aggregated replication metrics across fixed exits.
results/figures/plot_pareto.py Adds a script to plot a Pareto frontier from the fixed-exit metrics.
new_results/replication/opt_java/eval_exit_3.json Adds replication metrics for exit 3 (new_results).
new_results/replication/opt_java/eval_exit_5.json Adds replication metrics for exit 5 (new_results).
new_results/replication/opt_java/eval_exit_7.json Adds replication metrics for exit 7 (new_results).
new_results/replication/opt_java/eval_exit_9.json Adds replication metrics for exit 9 (new_results).
new_results/replication/opt_java/eval_exit_11.json Adds replication metrics for exit 11 (new_results).
new_results/replication/opt_java/eval_exit_13.json Adds replication metrics for exit 13 (new_results).
new_results/replication/opt_java/eval_exit_15.json Adds replication metrics for exit 15 (new_results).
new_results/replication/opt_java/eval_exit_19.json Adds replication metrics for exit 19 (new_results).
new_results/replication/opt_java/eval_exit_23.json Adds replication metrics for exit 23 (new_results).
new_results/replication/opt_java/eval_exit_27.json Adds replication metrics for exit 27 (new_results).
new_results/replication/opt_java/eval_exit_31.json Adds replication metrics for exit 31 (new_results).
new_results/replication/opt_java/eval_all_fixed_exits.json Adds aggregated replication metrics across fixed exits (new_results).
new_results/figures/plot_pareto.py Adds a script to plot a Pareto frontier from the fixed-exit metrics (new_results).

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread src/eval/evaluator.py
from zeus.monitor import ZeusMonitor

from src.models.opt.modeling_opt_single_head_fixed_exit import OPTFixedEESingleHead
from src.models.opt.modeling_opt_single_head_fixed_exit import OPTEESingleHead as OPTFixedEESingleHead
plt.grid(True, linestyle='--', alpha=0.7)
plt.plot(energy[-1], codebleu[-1], marker='*', color='gold', markersize=15, markeredgecolor='black')
plt.tight_layout()
out_path = os.path.expanduser('~/greencode/results/figures/opt_java_metrics_comparison_ctx_0.2.pdf')
plt.grid(True, linestyle='--', alpha=0.7)
plt.plot(energy[-1], codebleu[-1], marker='*', color='gold', markersize=15, markeredgecolor='black')
plt.tight_layout()
out_path = os.path.expanduser('~/greencode/results/figures/opt_java_metrics_comparison_ctx_0.2.pdf')
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants