[00:47:56] === sweep start ===
[00:47:56] RUN  gemma2-9b | medial | geo | idiom  -> results/sweep/json/gemma2-9b__medial__geo__idiom.json
[00:51:15]   OK  results/sweep/json/gemma2-9b__medial__geo__idiom.json
[00:51:15] RUN  gemma2-9b | medial | geo | nonidiom  -> results/sweep/json/gemma2-9b__medial__geo__nonidiom.json
[00:52:18]   OK  results/sweep/json/gemma2-9b__medial__geo__nonidiom.json
[00:52:18]   analyze -> results/sweep/reports/gemma2-9b__medial__geo__analyze.txt
Wrote report results/sweep/reports/gemma2-9b__medial__geo.md (6691 chars)
[00:52:20]   report  -> results/sweep/reports/gemma2-9b__medial__geo.md
Traceback (most recent call last):
  File "/home/prada/PID_evaluation/code/plot_hcov.py", line 21, in <module>
    import matplotlib
ModuleNotFoundError: No module named 'matplotlib'
[00:52:20] RUN  gemma2-9b | full | geo | idiom  -> results/sweep/json/gemma2-9b__full__geo__idiom.json
[00:55:22]   OK  results/sweep/json/gemma2-9b__full__geo__idiom.json
[00:55:22] RUN  gemma2-9b | full | geo | nonidiom  -> results/sweep/json/gemma2-9b__full__geo__nonidiom.json
[00:58:16]   OK  results/sweep/json/gemma2-9b__full__geo__nonidiom.json
[00:58:17]   analyze -> results/sweep/reports/gemma2-9b__full__geo__analyze.txt
Wrote report results/sweep/reports/gemma2-9b__full__geo.md (6775 chars)
[00:58:18]   report  -> results/sweep/reports/gemma2-9b__full__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [gemma2-9b__full__geo__idiom.json::ratio_s_idiom] dropped 9/18 non-finite values
  [gemma2-9b__full__geo__nonidiom.json::ratio_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [gemma2-9b__full__geo__idiom.json::h_s_idiom] dropped 9/18 non-finite values
  [gemma2-9b__full__geo__nonidiom.json::h_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/gemma2-9b__full__geo
[00:58:19]   plots   -> results/sweep/figures/gemma2-9b__full__geo
[00:58:19] RUN  gemma2-9b | medial | joint | idiom  -> results/sweep/json/gemma2-9b__medial__joint__idiom.json
[00:59:29]   OK  results/sweep/json/gemma2-9b__medial__joint__idiom.json
[00:59:29] RUN  gemma2-9b | medial | joint | nonidiom  -> results/sweep/json/gemma2-9b__medial__joint__nonidiom.json
[01:00:32]   OK  results/sweep/json/gemma2-9b__medial__joint__nonidiom.json
[01:00:32]   analyze -> results/sweep/reports/gemma2-9b__medial__joint__analyze.txt
Wrote report results/sweep/reports/gemma2-9b__medial__joint.md (6773 chars)
[01:00:33]   report  -> results/sweep/reports/gemma2-9b__medial__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [gemma2-9b__medial__joint__idiom.json::ratio_s_idiom] dropped 8/18 non-finite values
  [gemma2-9b__medial__joint__nonidiom.json::ratio_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [gemma2-9b__medial__joint__idiom.json::h_s_idiom] dropped 8/18 non-finite values
  [gemma2-9b__medial__joint__nonidiom.json::h_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/gemma2-9b__medial__joint
[01:00:35]   plots   -> results/sweep/figures/gemma2-9b__medial__joint
[01:00:35] RUN  gemma2-9b | full | joint | idiom  -> results/sweep/json/gemma2-9b__full__joint__idiom.json
[01:03:35]   OK  results/sweep/json/gemma2-9b__full__joint__idiom.json
[01:03:35] RUN  gemma2-9b | full | joint | nonidiom  -> results/sweep/json/gemma2-9b__full__joint__nonidiom.json
[01:06:34]   OK  results/sweep/json/gemma2-9b__full__joint__nonidiom.json
[01:06:35]   analyze -> results/sweep/reports/gemma2-9b__full__joint__analyze.txt
Wrote report results/sweep/reports/gemma2-9b__full__joint.md (6851 chars)
[01:06:36]   report  -> results/sweep/reports/gemma2-9b__full__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [gemma2-9b__full__joint__idiom.json::ratio_s_idiom] dropped 15/18 non-finite values
  [gemma2-9b__full__joint__nonidiom.json::ratio_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [gemma2-9b__full__joint__idiom.json::h_s_idiom] dropped 15/18 non-finite values
  [gemma2-9b__full__joint__nonidiom.json::h_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/gemma2-9b__full__joint
[01:06:37]   plots   -> results/sweep/figures/gemma2-9b__full__joint
[01:06:37] RUN  qwen3-8b-base | medial | geo | idiom  -> results/sweep/json/qwen3-8b-base__medial__geo__idiom.json
[01:08:13]   OK  results/sweep/json/qwen3-8b-base__medial__geo__idiom.json
[01:08:13] RUN  qwen3-8b-base | medial | geo | nonidiom  -> results/sweep/json/qwen3-8b-base__medial__geo__nonidiom.json
[01:08:53]   OK  results/sweep/json/qwen3-8b-base__medial__geo__nonidiom.json
[01:08:53]   analyze -> results/sweep/reports/qwen3-8b-base__medial__geo__analyze.txt
Wrote report results/sweep/reports/qwen3-8b-base__medial__geo.md (6711 chars)
[01:08:54]   report  -> results/sweep/reports/qwen3-8b-base__medial__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b-base__medial__geo__idiom.json::ratio_s_idiom] dropped 3/18 non-finite values
  [qwen3-8b-base__medial__geo__nonidiom.json::ratio_s_idiom] dropped 14/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b-base__medial__geo__idiom.json::h_s_idiom] dropped 3/18 non-finite values
  [qwen3-8b-base__medial__geo__nonidiom.json::h_s_idiom] dropped 14/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/qwen3-8b-base__medial__geo
[01:08:56]   plots   -> results/sweep/figures/qwen3-8b-base__medial__geo
[01:08:56] RUN  qwen3-8b-base | full | geo | idiom  -> results/sweep/json/qwen3-8b-base__full__geo__idiom.json
[01:11:04]   OK  results/sweep/json/qwen3-8b-base__full__geo__idiom.json
[01:11:04] RUN  qwen3-8b-base | full | geo | nonidiom  -> results/sweep/json/qwen3-8b-base__full__geo__nonidiom.json
[01:13:14]   OK  results/sweep/json/qwen3-8b-base__full__geo__nonidiom.json
[01:13:14]   analyze -> results/sweep/reports/qwen3-8b-base__full__geo__analyze.txt
Wrote report results/sweep/reports/qwen3-8b-base__full__geo.md (6621 chars)
[01:13:15]   report  -> results/sweep/reports/qwen3-8b-base__full__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b-base__full__geo__idiom.json::ratio_s_idiom] dropped 9/18 non-finite values
  [qwen3-8b-base__full__geo__nonidiom.json::ratio_s_idiom] dropped 18/18 non-finite values
  skipping ratio_s: no finite values (idioms=9, non-idioms=0)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b-base__full__geo__idiom.json::h_s_idiom] dropped 9/18 non-finite values
  [qwen3-8b-base__full__geo__nonidiom.json::h_s_idiom] dropped 18/18 non-finite values
  skipping h_s: no finite values (idioms=9, non-idioms=0)
Wrote 6 figures to results/sweep/figures/qwen3-8b-base__full__geo
[01:13:16]   plots   -> results/sweep/figures/qwen3-8b-base__full__geo
[01:13:16] RUN  qwen3-8b-base | medial | joint | idiom  -> results/sweep/json/qwen3-8b-base__medial__joint__idiom.json
[01:13:57]   OK  results/sweep/json/qwen3-8b-base__medial__joint__idiom.json
[01:13:57] RUN  qwen3-8b-base | medial | joint | nonidiom  -> results/sweep/json/qwen3-8b-base__medial__joint__nonidiom.json
[01:14:38]   OK  results/sweep/json/qwen3-8b-base__medial__joint__nonidiom.json
[01:14:38]   analyze -> results/sweep/reports/qwen3-8b-base__medial__joint__analyze.txt
Wrote report results/sweep/reports/qwen3-8b-base__medial__joint.md (6606 chars)
[01:14:39]   report  -> results/sweep/reports/qwen3-8b-base__medial__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b-base__medial__joint__idiom.json::ratio_s_idiom] dropped 10/18 non-finite values
  [qwen3-8b-base__medial__joint__nonidiom.json::ratio_s_idiom] dropped 18/18 non-finite values
  skipping ratio_s: no finite values (idioms=8, non-idioms=0)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b-base__medial__joint__idiom.json::h_s_idiom] dropped 10/18 non-finite values
  [qwen3-8b-base__medial__joint__nonidiom.json::h_s_idiom] dropped 18/18 non-finite values
  skipping h_s: no finite values (idioms=8, non-idioms=0)
Wrote 6 figures to results/sweep/figures/qwen3-8b-base__medial__joint
[01:14:41]   plots   -> results/sweep/figures/qwen3-8b-base__medial__joint
[01:14:41] RUN  qwen3-8b-base | full | joint | idiom  -> results/sweep/json/qwen3-8b-base__full__joint__idiom.json
[01:16:50]   OK  results/sweep/json/qwen3-8b-base__full__joint__idiom.json
[01:16:50] RUN  qwen3-8b-base | full | joint | nonidiom  -> results/sweep/json/qwen3-8b-base__full__joint__nonidiom.json
[01:18:56]   OK  results/sweep/json/qwen3-8b-base__full__joint__nonidiom.json
[01:18:57]   analyze -> results/sweep/reports/qwen3-8b-base__full__joint__analyze.txt
Wrote report results/sweep/reports/qwen3-8b-base__full__joint.md (6700 chars)
[01:18:58]   report  -> results/sweep/reports/qwen3-8b-base__full__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b-base__full__joint__idiom.json::ratio_s_idiom] dropped 13/18 non-finite values
  [qwen3-8b-base__full__joint__nonidiom.json::ratio_s_idiom] dropped 18/18 non-finite values
  skipping ratio_s: no finite values (idioms=5, non-idioms=0)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b-base__full__joint__idiom.json::h_s_idiom] dropped 13/18 non-finite values
  [qwen3-8b-base__full__joint__nonidiom.json::h_s_idiom] dropped 18/18 non-finite values
  skipping h_s: no finite values (idioms=5, non-idioms=0)
Wrote 6 figures to results/sweep/figures/qwen3-8b-base__full__joint
[01:18:59]   plots   -> results/sweep/figures/qwen3-8b-base__full__joint
[01:18:59] RUN  qwen3-8b | medial | geo | idiom  -> results/sweep/json/qwen3-8b__medial__geo__idiom.json
[01:19:49]   OK  results/sweep/json/qwen3-8b__medial__geo__idiom.json
[01:19:49] RUN  qwen3-8b | medial | geo | nonidiom  -> results/sweep/json/qwen3-8b__medial__geo__nonidiom.json
[01:20:30]   OK  results/sweep/json/qwen3-8b__medial__geo__nonidiom.json
[01:20:31]   analyze -> results/sweep/reports/qwen3-8b__medial__geo__analyze.txt
Wrote report results/sweep/reports/qwen3-8b__medial__geo.md (6704 chars)
[01:20:32]   report  -> results/sweep/reports/qwen3-8b__medial__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b__medial__geo__idiom.json::ratio_s_idiom] dropped 3/18 non-finite values
  [qwen3-8b__medial__geo__nonidiom.json::ratio_s_idiom] dropped 12/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b__medial__geo__idiom.json::h_s_idiom] dropped 3/18 non-finite values
  [qwen3-8b__medial__geo__nonidiom.json::h_s_idiom] dropped 12/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/qwen3-8b__medial__geo
[01:20:34]   plots   -> results/sweep/figures/qwen3-8b__medial__geo
[01:20:34] RUN  qwen3-8b | full | geo | idiom  -> results/sweep/json/qwen3-8b__full__geo__idiom.json
[01:22:42]   OK  results/sweep/json/qwen3-8b__full__geo__idiom.json
[01:22:42] RUN  qwen3-8b | full | geo | nonidiom  -> results/sweep/json/qwen3-8b__full__geo__nonidiom.json
[01:24:49]   OK  results/sweep/json/qwen3-8b__full__geo__nonidiom.json
[01:24:50]   analyze -> results/sweep/reports/qwen3-8b__full__geo__analyze.txt
Wrote report results/sweep/reports/qwen3-8b__full__geo.md (6589 chars)
[01:24:51]   report  -> results/sweep/reports/qwen3-8b__full__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b__full__geo__idiom.json::ratio_s_idiom] dropped 14/18 non-finite values
  [qwen3-8b__full__geo__nonidiom.json::ratio_s_idiom] dropped 18/18 non-finite values
  skipping ratio_s: no finite values (idioms=4, non-idioms=0)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b__full__geo__idiom.json::h_s_idiom] dropped 14/18 non-finite values
  [qwen3-8b__full__geo__nonidiom.json::h_s_idiom] dropped 18/18 non-finite values
  skipping h_s: no finite values (idioms=4, non-idioms=0)
Wrote 6 figures to results/sweep/figures/qwen3-8b__full__geo
[01:24:52]   plots   -> results/sweep/figures/qwen3-8b__full__geo
[01:24:52] RUN  qwen3-8b | medial | joint | idiom  -> results/sweep/json/qwen3-8b__medial__joint__idiom.json
[01:25:32]   OK  results/sweep/json/qwen3-8b__medial__joint__idiom.json
[01:25:32] RUN  qwen3-8b | medial | joint | nonidiom  -> results/sweep/json/qwen3-8b__medial__joint__nonidiom.json
[01:26:13]   OK  results/sweep/json/qwen3-8b__medial__joint__nonidiom.json
[01:26:13]   analyze -> results/sweep/reports/qwen3-8b__medial__joint__analyze.txt
Wrote report results/sweep/reports/qwen3-8b__medial__joint.md (6761 chars)
[01:26:15]   report  -> results/sweep/reports/qwen3-8b__medial__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b__medial__joint__idiom.json::ratio_s_idiom] dropped 9/18 non-finite values
  [qwen3-8b__medial__joint__nonidiom.json::ratio_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b__medial__joint__idiom.json::h_s_idiom] dropped 9/18 non-finite values
  [qwen3-8b__medial__joint__nonidiom.json::h_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/qwen3-8b__medial__joint
[01:26:16]   plots   -> results/sweep/figures/qwen3-8b__medial__joint
[01:26:16] RUN  qwen3-8b | full | joint | idiom  -> results/sweep/json/qwen3-8b__full__joint__idiom.json
[01:28:26]   OK  results/sweep/json/qwen3-8b__full__joint__idiom.json
[01:28:26] RUN  qwen3-8b | full | joint | nonidiom  -> results/sweep/json/qwen3-8b__full__joint__nonidiom.json
[01:30:34]   OK  results/sweep/json/qwen3-8b__full__joint__nonidiom.json
[01:30:34]   analyze -> results/sweep/reports/qwen3-8b__full__joint__analyze.txt
Wrote report results/sweep/reports/qwen3-8b__full__joint.md (6671 chars)
[01:30:35]   report  -> results/sweep/reports/qwen3-8b__full__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [qwen3-8b__full__joint__idiom.json::ratio_s_idiom] dropped 16/18 non-finite values
  [qwen3-8b__full__joint__nonidiom.json::ratio_s_idiom] dropped 18/18 non-finite values
  skipping ratio_s: no finite values (idioms=2, non-idioms=0)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [qwen3-8b__full__joint__idiom.json::h_s_idiom] dropped 16/18 non-finite values
  [qwen3-8b__full__joint__nonidiom.json::h_s_idiom] dropped 18/18 non-finite values
  skipping h_s: no finite values (idioms=2, non-idioms=0)
Wrote 6 figures to results/sweep/figures/qwen3-8b__full__joint
[01:30:36]   plots   -> results/sweep/figures/qwen3-8b__full__joint
[01:30:36] SKIP (exists) results/sweep/json/llama3.1-8b__medial__geo__idiom.json
[01:30:36] RUN  llama3.1-8b | medial | geo | nonidiom  -> results/sweep/json/llama3.1-8b__medial__geo__nonidiom.json
[01:31:21]   OK  results/sweep/json/llama3.1-8b__medial__geo__nonidiom.json
[01:31:21]   analyze -> results/sweep/reports/llama3.1-8b__medial__geo__analyze.txt
Wrote report results/sweep/reports/llama3.1-8b__medial__geo.md (6731 chars)
[01:31:23]   report  -> results/sweep/reports/llama3.1-8b__medial__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [llama3.1-8b__medial__geo__idiom.json::ratio_s_idiom] dropped 1/18 non-finite values
  [llama3.1-8b__medial__geo__nonidiom.json::ratio_s_idiom] dropped 13/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [llama3.1-8b__medial__geo__idiom.json::h_s_idiom] dropped 1/18 non-finite values
  [llama3.1-8b__medial__geo__nonidiom.json::h_s_idiom] dropped 13/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/llama3.1-8b__medial__geo
[01:31:24]   plots   -> results/sweep/figures/llama3.1-8b__medial__geo
[01:31:24] RUN  llama3.1-8b | full | geo | idiom  -> results/sweep/json/llama3.1-8b__full__geo__idiom.json
[01:33:07]   OK  results/sweep/json/llama3.1-8b__full__geo__idiom.json
[01:33:07] RUN  llama3.1-8b | full | geo | nonidiom  -> results/sweep/json/llama3.1-8b__full__geo__nonidiom.json
[01:34:50]   OK  results/sweep/json/llama3.1-8b__full__geo__nonidiom.json
[01:34:50]   analyze -> results/sweep/reports/llama3.1-8b__full__geo__analyze.txt
Wrote report results/sweep/reports/llama3.1-8b__full__geo.md (6788 chars)
[01:34:52]   report  -> results/sweep/reports/llama3.1-8b__full__geo.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [llama3.1-8b__full__geo__idiom.json::ratio_s_idiom] dropped 10/18 non-finite values
  [llama3.1-8b__full__geo__nonidiom.json::ratio_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [llama3.1-8b__full__geo__idiom.json::h_s_idiom] dropped 10/18 non-finite values
  [llama3.1-8b__full__geo__nonidiom.json::h_s_idiom] dropped 17/18 non-finite values
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:219: RuntimeWarning: Degrees of freedom <= 0 for slice
  ret = _var(a, axis=axis, dtype=dtype, out=out, ddof=ddof,
/home/prada/PID_evaluation/.venv/lib/python3.13/site-packages/numpy/_core/_methods.py:211: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/llama3.1-8b__full__geo
[01:34:53]   plots   -> results/sweep/figures/llama3.1-8b__full__geo
[01:34:53] RUN  llama3.1-8b | medial | joint | idiom  -> results/sweep/json/llama3.1-8b__medial__joint__idiom.json
[01:35:27]   OK  results/sweep/json/llama3.1-8b__medial__joint__idiom.json
[01:35:27] RUN  llama3.1-8b | medial | joint | nonidiom  -> results/sweep/json/llama3.1-8b__medial__joint__nonidiom.json
[01:36:01]   OK  results/sweep/json/llama3.1-8b__medial__joint__nonidiom.json
[01:36:01]   analyze -> results/sweep/reports/llama3.1-8b__medial__joint__analyze.txt
Wrote report results/sweep/reports/llama3.1-8b__medial__joint.md (6818 chars)
[01:36:03]   report  -> results/sweep/reports/llama3.1-8b__medial__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [llama3.1-8b__medial__joint__idiom.json::ratio_s_idiom] dropped 3/18 non-finite values
  [llama3.1-8b__medial__joint__nonidiom.json::ratio_s_idiom] dropped 16/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [llama3.1-8b__medial__joint__idiom.json::h_s_idiom] dropped 3/18 non-finite values
  [llama3.1-8b__medial__joint__nonidiom.json::h_s_idiom] dropped 16/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/llama3.1-8b__medial__joint
[01:36:04]   plots   -> results/sweep/figures/llama3.1-8b__medial__joint
[01:36:04] RUN  llama3.1-8b | full | joint | idiom  -> results/sweep/json/llama3.1-8b__full__joint__idiom.json
[01:37:47]   OK  results/sweep/json/llama3.1-8b__full__joint__idiom.json
[01:37:47] RUN  llama3.1-8b | full | joint | nonidiom  -> results/sweep/json/llama3.1-8b__full__joint__nonidiom.json
[01:39:30]   OK  results/sweep/json/llama3.1-8b__full__joint__nonidiom.json
[01:39:30]   analyze -> results/sweep/reports/llama3.1-8b__full__joint__analyze.txt
Wrote report results/sweep/reports/llama3.1-8b__full__joint.md (6894 chars)
[01:39:31]   report  -> results/sweep/reports/llama3.1-8b__full__joint.md
metric ratio_u (ratio_u_idiom):
metric ratio_s (ratio_s_idiom):
  [llama3.1-8b__full__joint__idiom.json::ratio_s_idiom] dropped 10/18 non-finite values
  [llama3.1-8b__full__joint__nonidiom.json::ratio_s_idiom] dropped 16/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_u (h_u_idiom):
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
metric h_s (h_s_idiom):
  [llama3.1-8b__full__joint__idiom.json::h_s_idiom] dropped 10/18 non-finite values
  [llama3.1-8b__full__joint__nonidiom.json::h_s_idiom] dropped 16/18 non-finite values
/home/prada/PID_evaluation/code/plot_hcov.py:131: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="upper right")
/home/prada/PID_evaluation/code/plot_hcov.py:156: UserWarning: No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.
  ax.legend(loc="lower right", fontsize=8)
Wrote 12 figures to results/sweep/figures/llama3.1-8b__full__joint
[01:39:33]   plots   -> results/sweep/figures/llama3.1-8b__full__joint
[01:39:33] === building master summary ===
Wrote summary results/sweep/SUMMARY.md (16 configs)
[01:39:42] summary -> results/sweep/SUMMARY.md
[01:39:42] === sweep done ===